RFP STATE 0000000538SL - Career Readiness Assessments.docx

State of Missouri Office of Administration, Division of Purchasing

Request for Proposal (RFP) for Career Readiness, Assessments, and Certifications Web-Based Platform

PROPOSAL DUE NO LATER THAN: SEPTEMBER 10, 2026

AT 2:00 PM CENTRAL TIME

Proposal response must be submitted electronically through MissouriBUYS, powered by MOVERS, at https://missouribuys.mo.gov.

E-mailed, mailed, courier, or hand-delivered proposal responses will not be accepted.

RFP CONTACT INFORMATION:

BUYER: Madison Lunte

PHONE NO.: (573) 751-8058

EMAIL: Madison.Lunte@oa.mo.gov See “RFP Questions” in Section 1 for appropriate communications during the procurement process.

DELIVER SUPPLIES/SERVICES FREE ON BOARD (FOB) DESTINATION TO THE FOLLOWING ADDRESS:

Department of Higher Education and Workforce Development 301 West High Street, Suite 840 Jefferson City, MO 65101

ATTENTION:

1. After reviewing the Request for Proposal (RFP), the vendor must complete and return Exhibit A, Proposal Signature Page and all other necessary exhibits.

2. Due to lead times for obtaining the information needed to complete the various Business Compliance Exhibits herein as explained in the RFP’s Vendor Response Exhibits, vendors are encouraged to IMMEDIATELY begin securing these verifications.

3. The vendor must be registered in MissouriBUYS, powered by MOVERS in a “Prospective” or “Spend Authorized” registration status to submit a proposal. The vendor must achieve “Approved” registration status in MissouriBUYS (WebProcure/Proactis) and “Spend Authorized” registration status in MissouriBUYS, powered by MOVERS to be considered for a contract award. Reference Section 5.

RFP Organization:

RFP Vendor Response Exhibits (Return these exhibits with the proposal)

BUSINESS COMPLIANCE EXHIBITS

Separate Documents: The vendor is advised that a separate document to this document referenced above provide additional requirements, information, and/or instruction. The separate document must be downloaded from the Division of Purchasing’s MissouriBUYS, powered by MOVERS, website at: https://missouribuys.mo.gov/. The separate document is downloadable from the same web page where the solicitation document is downloadable. It shall be the sole responsibility of the vendor to obtain the separate document. The vendor shall not be relieved of any responsibility for performance under the subsequent contract due to the failure of the vendor to obtain a copy of the separate document.

INTRODUCTION AND BACKGROUND INFORMATION SECTION

Introduction:

Purpose: This document constitutes a request for competitive, sealed proposals for the provision of Career Readiness, Assessments and Certifications Web-Based Platform with assessments, National Career Readiness Certificates (NCRC) or equivalent, for the purpose of assessing the skills of Missouri’s workforce, that includes curriculum for essential skills and career readiness skill, as set forth herein for the Department of Higher Education and Workforce Development (DHEWD) (hereinafter referred to as “state agency”).

Minimum Experience Qualifications: There are minimum experience qualifications in the RFP (See Section 2.2, Minimum Experience Requirements in Section 2: Scope of Work).

Titles: Titles of paragraphs used herein are for the purpose of facilitating reference only and shall not be construed to infer a contractual construction of language.

Background and Historical Usage Information:

DHEWD has utilized the ACT WorkKeys products since 2004 to meet their National Career Readiness Certification (NCRC) initiative. For the past twenty-two (22) years, through the Missouri Job Centers, the Office of Workforce Development (OWD) has provided this resource statewide to employers and job seekers to measure skills that are essential to succeed in the workplace. These assessments demonstrate the job seekers’ strengths and weaknesses and provide a valid way to show job seekers abilities to employers. Missouri currently has 112 counties participating in this initiative with 99 counties having received certification.

Current Contract Information:

A current contract exists for the services being obtained via this RFP. A copy of the contracts can be viewed and printed from the Division of Purchasing Awarded Bid & Contract Document Search System located on the Internet at: https://purch.oa.mo.gov/bidding-contracts/awarded-bid-contract-document-search. In addition, all proposal and evaluation documentation leading to the award of that contract may also be viewed and printed from the Division of Purchasing’s Awarded Bid & Contract Document Search System. Please reference the Single Feasible Source number SFST300349021017385 or the contract number CT211783002 when searching for these documents.

RFP Questions:

Buyer is Single Point of Contact for Solicitation: Vendors and their agents (including subcontractors, employees, consultants, or anyone else acting on their behalf) must direct all of their questions or comments regarding the RFP, the evaluation, etc., to the buyer indicated on the first page of this RFP. It is preferred that questions be emailed to the buyer.

Except as noted herein, vendors and their agents are instructed not to contact any other state employee regarding any of these matters during the solicitation and evaluation process. Inappropriate contacts are grounds for suspension and/or exclusion from specific procurements. Vendors can be sanctioned for unauthorized contact with any evaluator under 1 Code of State Regulation (CSR) 40-1.060(8)(G) and (H) available at http://www.sos.mo.gov/adrules/csr/csr.asp.

Service-Disabled Veteran Enterprise certification or subcontract questions should be directed to the Office of Equal Opportunity (OEO) at http://oeo.mo.gov.

Vendor is Responsible for Asking Questions About the RFP: It is the vendor’s responsibility to ask questions, request changes or clarifications, or otherwise advise the Division of Purchasing if the vendor believes that any language, specifications, or requirements are: (1) ambiguous, (2) contradictory or arbitrary, (3) violate any state or federal law or regulation, (4) restrict or limit the requirements to a single source, or (5) restrict or limit the vendor’s ability to submit a proposal.

Vendor Question Deadline: Every attempt shall be made to ensure that the vendor receives an adequate and prompt response. However, in order to maintain a fair and equitable procurement process, all vendors will be advised, via the issuance of an amendment to the RFP, of any relevant or pertinent information related to the procurement. All questions and issues should be submitted no later than ten (10) calendar days prior to the proposal closing date and time of the proposals. If not received prior to ten (10) calendar days before the proposal closing date and time, the Division of Purchasing may not be able to fully research and consider the respective questions or issues.

State’s Response to Vendor Questions: Upon the Division of Purchasing’s consideration of questions and issues, if the Division of Purchasing determines that changes are necessary, the resulting changes will be included in a subsequently issued RFP amendment(s); absence of such response indicates that the questions and issues were considered but deemed unnecessary for an RFP amendment. All vendors will be advised of any change to the RFP’s language, specifications, or requirements by a formal amendment to the RFP. There will be no posted written records of the questions/communications (i.e., formal question/answer document).

RFP is State’s Only Official Position: The only official position of the State of Missouri shall be that which is contained in the RFP and any amendments thereto.

Amendments:

If the Division of Purchasing determines that changes to the RFP are necessary, the resulting changes will be included in a subsequently issued RFP amendment(s) prior to the proposal closing date and time.

Pre-Proposal Teleconference:

Pre-proposal Teleconference Date and Time: A pre-proposal teleconference regarding this RFP will be held on Monday August 17, 2026 at 3:00 p.m., Central Time. Vendors are encouraged to participate in the pre-proposal teleconference as it will be used as the forum for questions, communications, and discussions regarding the RFP. The vendor should become familiar with the RFP and develop all questions prior to the pre-proposal teleconference in order to ask questions and otherwise participate in the public communications regarding the RFP.

Pre-Proposal Teleconference Agenda: The vendor should have a copy of the RFP for the pre-proposal teleconference since it will be used as the agenda for the pre-proposal teleconference.

Communication Prior to the Teleconference: The vendor is encouraged to review the RFP prior to the pre-proposal teleconference and may submit written communications and/or questions regarding the RFP, which reference the RFP paragraph numbers, to the buyer identified on page one. Such prior communication will provide the State of Missouri with insight into areas of the RFP which may be brought up for discussion during the conference and which may require clarification.

Communication During the Teleconference: During the pre-proposal teleconference, it shall be the sole responsibility of the vendor to orally address all issues previously presented to the buyer by the vendor, including any questions regarding the RFP or areas of the RFP requiring clarification.

Amendment to the RFP: Any modifications needed to the RFP as a result of discussions from the pre-proposal teleconference will be accomplished as an amendment to the RFP. Neither formal minutes of the pre-proposal teleconference nor written records of the questions/communications will be maintained.

Pre-Proposal Teleconference Special Accommodations: Vendors are strongly encouraged to advise the Division of Purchasing within five (5) state business days of the scheduled pre-proposal teleconference of any special accommodations needed for people with health conditions or impairments who will be participating in the pre-proposal teleconference so that these accommodations can be made.

Pre-Proposal Teleconference Dial-In Information: The vendor should contact the buyer from the Division of Purchasing as indicated on the first page of this RFP to obtain dial-in instructions. The vendor will be provided with a telephone number to dial and/or webinar information to access, in order to listen and participate in the pre-proposal teleconference. The vendor shall refrain from contacting anyone other than the buyer to obtain the dial-in information.

Glossary of Terms and Acronyms:

Whenever the following terms and acronyms appear in the RFP document or any amendment thereto, the definitions or meanings described below shall apply.

General Glossary, Acronyms, and Abbreviations:

Administrator refers to individuals designated by the Department of Higher Education and Workforce Development to have overall access to the web-based platform and its capabilities.

Agency and/or State Agency means the statutory unit of state government in the State of Missouri for which the equipment, supplies, and/or services are being purchased by the Division of Purchasing (Purchasing). The agency is also responsible for payment, unless otherwise specified herein.

Amendment means a written, official modification to a solicitation or contract.

Attachment applies to all documents which are included with an RFP to incorporate any informational data or requirements related to the performance requirements and/or specifications.

Buyer means the procurement staff member of Purchasing.

Code of State Regulation (CSR) contains the current administrative rules of executive agencies of Missouri government. The regulations are arranged by agency rather than by subject.

Contract means a legal and binding agreement between two or more competent parties, for a consideration for the procurement of equipment, supplies, and/or services.

Contractor means a supplier, offeror, person, or organization who is a successful vendor as a result of an RFP and who enters into a contract.

Exhibit applies to forms which are included with an RFP for the vendor to complete and submit with their proposal prior to the specified closing date and time.

Individual participant refers to any person who is taking the assessment in order to earn a certificate. This includes any participant undergoing curriculum education prior to taking the assessment.

May means that a certain feature, component, or action is permissible, but not required.

Must means that a certain feature, component, or action is a mandatory condition.

Party refers to either the State of Missouri or the contractor as an entity that may enter into a contract pursuant to the terms herein.

Pricing Page(s) applies to the form(s) on which the vendor must state the price(s) applicable for the equipment, supplies, and/or services required in the RFP. The pricing pages must be completed and submitted by the vendor with the proposal prior to the specified proposal closing date and time.

Proctor refers to the individual who is administering the assessments to the individual participant and ensuring test compliance during the assessment.

Proposal Closing Date and Time and similar expressions mean the exact deadline required by the RFP for the receipt of sealed proposals.

Purchase Order means the authorized document issued by the state agency to the contractor indicating descriptions, quantities, and agreed prices for products and/or services.

Reasonable, Necessary or Proper as used herein shall be interpreted solely by the State of Missouri.

Request for Proposal (RFP) means the solicitation document issued by Purchasing to potential vendors for the purchase of equipment, supplies, and/or services as described in the document. The definition includes the following sections: Introduction and Background Information; Scope of Work; Terms and Conditions (“terms and conditions” and “Terms and Conditions” are used interchangeably throughout the RFP); General Contractual Requirements; and Vendor Submission, Evaluation, and Award Information; and the RFP Vendor Response Exhibits, Attachments, and Amendments of the RFP.

RSMo (Revised Statutes of Missouri) refers to the body of laws enacted by the Legislature which govern the operations of all agencies of the State of Missouri. Chapter 34 of the statutes is the primary chapter governing the operations of Purchasing.

Shall has the same meaning as the word must.

Should means that a certain feature, component and/or action is desirable but not mandatory.

State collectively referring to the state government and/or the agencies thereof.

Supplier has the same meaning as the word, vendor.

Vendor means the supplier, offeror, person, or organization that responds to an RFP by submitting a proposal with prices to provide the equipment, supplies, and/or services as required in the RFP document.

0. Artificial Intelligence (AI) Definitions:

0. Artificial Intelligence (AI): A machine-based system or service that can generate outputs such as content, predictions, recommendations, or decisions (i.e. virtual or real environments) influencing the environments they interact with.

0. Generative AI: Artificial intelligence (AI) service that generates new content, such as text, images, audio, code, synthetic data, or other content based on learned patterns from large data sets, including but not limited to Large Language Models (LLMs).

0. Machine Learning (ML): A type of Artificial Intelligence (AI) that helps computers and machines learn like humans, to perform tasks, and to improve their performance and accuracy through experience and exposure to more data.

0. AI Training: The computational process by which an artificial intelligence or machine learning model ingests and processes data to develop, adjust, or optimize its internal parameters, weights, or algorithms for the purpose of enabling the model to perform specified tasks, generate outputs, or make predictions. Training includes initial pre-training on large datasets, fine-tuning on domain-specific data, and any reinforcement or feedback-based optimization processes applied to the model.

0. Automated Decision-Making (ADM) Technology: Any system, software, or algorithmic process—utilizing either artificial intelligence, machine learning, or advanced rule-based logic—that automates the evaluation of data to make a decision, issue a score, or recommend a specific course of action with minimal or no human intervention.

0. AI Model: A component of an information system that implements AI technology and uses computational, statistical, or machine-learning techniques to produce outputs from a given set of inputs.

0. AI Prompt: The input, instruction, or command that a user provides to an Artificial Intelligence (AI) model to elicit a specific AI generated output, which can be in the form of text, code, images, audio, automated task or function.

0. Agentic AI: Artificial Intelligence (AI) service that exhibits a level of autonomy, proactivity, and decision-making capability, often acting as agents that can pursue goals, initiate actions, and adapt based on context without direct human prompting.

0. Data Poisoning: A type of cyberattack in which an adversary intentionally compromises a training dataset used by an AI or Machine Learning (ML) model to influence or manipulate the operation of that model.

0. Drift: A phenomenon where an AI model's performance degrades over time due to changes in the data it is exposed to, causing its predictions or outputs to become less accurate or relevant.

0. Hallucination: When an Artificial Intelligence (AI) service generates inaccurate or nonsensical information.

0. Model Bias: When an Artificial Intelligence (AI) service makes systematic errors or shows favor to certain groups, leading to unfair or discriminatory outcomes.

0. Prompt Injection: An attack that forces a generative model to produce unexpected output by manipulating the structure, instructions, or information contained in its prompt.

0. Deterministic AI: A machine-based system that operates on predefined rules to produce consistent, predictable outputs for a given input, without the variability typically found in generative models.

0. Open-Weight AI: Artificial intelligence systems where the underlying source code, model weights, and training data are made publicly accessible for review, use, and community development.

0. Closed AI: Proprietary artificial intelligence systems where the technology, code, training data, and model weights remain private and are strictly controlled by the developer or vendor.

0. AI Token: A basic unit of text that Generative Artificial Intelligence (AI) uses to understand and generate language. A token may be an entire word or parts of a word. This is often how vendors will charge you "by the token" for use of their AI services.

0. AI Tuning: The process of adjusting and refining an existing artificial intelligence model to make it more accurate, efficient, or skilled at a specific job.

0. Algorithm: An algorithm is simply a specific set of step-by-step instructions that a computer follows to solve a problem or complete a task. Common algorithms include classification, regression, and clustering.

0. Artificial Neural Network: A type of artificial intelligence (AI) designed with a deep learning technique modeled after the structure of the human brain, requiring large datasets to perform complex calculations and generate outputs - enabling capabilities such as speech and image recognition.

0. Chatbot: Software application that is designed to imitate human conversation through text or voice commands. See also Generative Artificial Intelligence or GenAI.

0. Deep Learning: A subfield of machine learning that focuses on training neural networks with many layers, enabling learning of complex patterns.

0. Domain Language Model: A domain language model is a type of artificial intelligence trained to understand and write about a highly specific subject, rather than general, everyday topics. For example, the state might use one specifically trained on Missouri legal codes or public health regulations so it can accurately help staff search specialized documents or answer complex policy questions.

0. Embedded AI: The Artificial Intelligence (AI) capabilities are part of an existing application as an additional feature of that application

0. Fine-tuned model: An Artificial Intelligence (AI) program that has already learned general information, but is then given extra training and is focused on a specific context or category of information so it becomes an expert in that area.

0. Input Transformation: A data processing technique where raw input is systematically altered, rephrased, or sanitized before being passed to a downstream system.

0. Large Language Model: A type of artificial intelligence model designed to process and generate human-like text by leveraging deep learning techniques, typically trained on vast amounts of data to understand and produce language-based outputs.

0. Local-side secret scanning: Is a security process where data is analyzed for sensitive information (passwords, API keys, tokens) on the user’s device before it is ever transmitted over the network. (sometimes known as client-side secret scanning)

0. Model Context Protocol: An open-source framework designed to standardize how Artificial Intelligence (AI) systems, particularly Large Language Models (LLMs), integrate with and share data with external tools, systems, and data sources.

0. Multimodal AI Model: An AI model capable of processing and integrating multiple data types - such as text, images, and audio - simultaneously. By combining information from these different modalities, the model can generate more accurate and context-aware responses to user queries or tasks.

0. Natural Language Processing: A term for a machine’s ability to perform conversational tasks, such as recognizing what is said to it, understanding the intended meaning and responding intelligibly.

0. Neural Network: AI technology mimics human brain function to process data and create patterns for decision making.

0. Non-Deterministic Noise: Refers to variability in system behavior that cannot be predicted or reproduced reliably from the same inputs. In coding, it manifests as outputs or states that differ across runs despite identical code, configuration, and initial conditions.

0. Power User (AI): A Power user (AI) primary use goes beyond the limits of the standard AI tool and uses it for deep research, coding, or analyzing massive datasets.

0. Pretrained Model: An Artificial Intelligence (AI) program that has already been taught to understand basic language, images, or patterns by analyzing massive amounts of general information. A pretrained model can be used as a starting point to create a fine-tuned contextualized version of a model, thus applying transfer learning.

0. Private AI Model: An Artificial Intelligence (AI) system that trained on the state's internal data and is used in a single tenant environment available only to the state.

0. Prompt Engineering: Prompt engineering is the practice of carefully writing the instructions or questions (the "prompts") given to an artificial intelligence tool to get the best and most accurate response. Think of it like giving highly specific, step-by-step directions to a new assistant so they understand exactly what you need them to do without getting confused.

0. Public AI Model: An Artificial Intelligence (AI) system trained on publicly available data from across the internet—such as text articles, images, and videos—and used in a multi-tenant environment, where it is accessible to multiple organizations. This means that, as a customer, your data may not be completely private, and some of these models may use your input to improve their algorithms.

0. Red Teaming: An exercise where human testers (or automated adversarial agents) attempt to force an AI model to produce harmful, illegal, or unethical outputs that violate its own safety guardrails.

0. Retrieval Augmented Generation: An AI technique for improving the quality of Generative AI (GenAI) responses by including trusted sources of knowledge, outside of the original training set, to improve the accuracy of the GenAI’s output. Implementing Retrieval Augmented Generation (RAG) in Generative AI question answering system has benefits: 1) assurance that the service has access to the most current, reliable facts, 2) reduce hallucinations rates, and 3) provide source attribution to increase user trust in the output.

0. Standard User (AI): A Standard User that uses AI for "General Productivity." They primarily use the service to help with administrative tasks, drafting emails, and summarizing documents.

0. Sub-Agent: An Artificial Intelligence (AI) agent that receives a bounded task from an orchestrator and executes it within defined constraints. Sub-agents may themselves further delegate to other agents.

0. Vibe Coding: A software development practice that shifts your role from writing code line-by-line to guiding an Artificial Intelligence (AI) assistant through a conversational process. Instead of manual syntax and precision, you describe what you want in plain language, and agentic AI systems turn that intent into working code.

Accuracy of Background Information:

Although an attempt has been made to provide accurate and up-to-date information, the State of Missouri does not warrant or represent that the background information provided herein reflects all relationships or existing conditions related to this RFP.

****END OF INTRODUCTION AND BACKGROUND INFORMATION SECTION****

SCOPE OF WORK SECTION

General Requirements:

The contractor shall provide a Career Readiness, Assessments and Certifications Web-Based Platform for the Department of Higher Education and Workforce Development (DHEWD) in accordance with the provisions and requirements stated herein and to the sole satisfaction of the state agency.

Contractor’s Obligation: Unless otherwise specified herein, the contractor shall furnish all material, labor, facilities, equipment, and supplies necessary to provide the products required herein.

The following various institutions are authorized to utilize services under the awarded contract:

The Office of Workforce Development (OWD): OWD may place an order for products and services listed in Exhibit C (Pricing Page).

Other entities that may receive pricing under contract and are included in the Missouri Office of Workforce Development (OWD) web-based platform include but are not limited to: publicly funded two-year colleges in Missouri.

Missouri Workforce Development Boards: These institutions may order individual products and services listed in Exhibit C (Pricing Page).

The State of Missouri provides no guarantee of quantities of products or services that may be acquired through the use of this contract.

Minimum Experience Requirements:

The contractor must meet or exceed the following minimum experience requirements at the time of proposal submission and for the duration of the contract:

The contractor must have a minimum of two (2) years of experience in issuing National Career Readiness Certificates (NCRC’s) or equivalent certificates.

The contractor must have a minimum of two (2) years of experience working with a government entity that is similar scope and size of the Department of Higher Education and Workforce Development.

The contractor must have a minimum of two (2) years of experience providing system maintenance (e.g. upgrades/new releases) and technical support for career readiness, assessments and certifications web-based platform-related software.

The contractor should have experience providing online curriculum for the following:

Essential workforce skills;

Primary assessments in applied math, graphic literacy and workplace documents; and Additional assessments on workplace observation and applied technology as needed.

The contractor should have experience in issuing electronic NCRC’s, or equivalent certificates, that are accepted by employers, to the clients, within a week of online testing.

Personnel Requirements

The contractor shall include the following personnel members to administer and perform the contract requirements:

Project Manager: The contractor’s project manager must have two (2) years of experience working with a state or public entity of comparable size to Missouri to provide a career readiness, assessments and certifications web-based platform.

Support Personnel: The contractor must provide an adequate number of support personnel to administer the contract and provide support to the state agency for the entirety of the contract.

Contact Person: The contractor shall designate a contact person who shall serve as the contractor’s contact and shall be the liaison between the contractor and the state agency. By no later than five (5) state business days after authorization by the state agency to proceed with services, the contractor shall provide the state agency with the name, address, email address, and telephone number of the contractor’s /contact person. The contractor’s contact person shall (1) oversee all services being provided, (2) assume responsibility and liability for services performed per the contract, and (3) serve as the primary point of contact with the state agency.

Background Checks: Prior to providing services, the contractor must ensure all personnel (including third-party staff) pass nationwide fingerprint-based background checks through the Missouri State Highway Patrol (https://www.machs.mo.gov/MACHSFP/home.html) that meet or exceed state standards for employees. The contractor shall be responsible for all associated costs. The state reserves the right to require evidence of compliance or a formal attestation regarding the results and the contractor’s background check process for audit purposes.

0. Artificial Intelligence (AI) Requirements:

1. The contractor shall disclose in writing whether any solution, system, service, or deliverable provided under this contract incorporates or relies upon Artificial Intelligence (AI), Machine Learning, or Automated Decision-Making Technologies. Such disclosure must occur prior to implementation or use in the performance of state business and must be approved by the state, in writing, before the contractor may proceed. The state reserves the right to request additional technical and risk documentation prior to granting approval. Failure to disclose such utilization shall constitute a material breach of contract.

0. Initial disclosure of the intent to use AI shall be provided by the contractor in their proposal response prior to contract execution.

0. If any AI, Machine Learning, or Automated Decision-Making Technology is introduced after initial disclosure prior to contract award, the contractor shall provide the state with a minimum of 20 business days' advance written notice to state designated contract manager, who will coordinate technical and risk review with the Information Technology Services Division, prior to utilizing any new, previously undisclosed, or third-party AI (including Generative AI), Machine Learning, or Automated Decision-Making Technologies in the performance of its obligations under this contract. The state retains sole discretion to approve or deny the contractor's use or integration of these technologies into the contract work. The contractor shall not proceed with the integration or use of these technologies unless and until written approval is granted by the state. Failure to respond by the state shall not be construed as approval.

1. If AI is used to develop or perform any services under the contract, the contractor must not transmit, process, analyze, store, train, fine-tune, or otherwise use any State Data in any Artificial Intelligence (AI) system—including Generative AI (GenAI), Large Language Models (LLMs), Machine Learning models, external analytics services, or Automated Decision-Making Technologies—without prior written authorization from the state.

1. Notwithstanding any authorization granted by the previous statement, the contractor shall not use State Data, or any derivative data, to train, fine-tune, or enhance any proprietary or third-party AI models.

1. The contractor shall ensure that its employees, agents, subcontractors, affiliates, and any third parties acting on its behalf strictly comply with all AI provisions set forth in this contract. The contractor shall be fully responsible and strictly liable for any acts, omissions, or violations of these requirements by such parties. The contractor shall agree to defend, indemnify, and hold harmless the state from any claims, damages, liabilities, or expenses (including legal fees) arising out of or related to the unauthorized or non-compliant use of AI technologies by the contractor or any party acting on its behalf.

1. The state reserves the right to deny the use of any AI system for any reason, or to rescind previous approval, at its sole discretion.

1. The state reserves the right to amend the contract, without additional cost, to incorporate updated AI terms, conditions, and security protocols. Furthermore, the state may cancel this Contract immediately, without penalty, if it determines in its sole discretion that the contractor’s use or integration of AI, Machine Learning, or Automated Decision Making Technologies, pose unacceptable risk(s) to state data, operations, or cybersecurity.

1. If AI is used to provide the system or services required herein, the contractor must comply with NIST AI Risk Management Framework 1.0, or newer versions, as required to satisfy an audit from a federal agency, state agency, or other regulatory body, including framework revisions adopted by these entities.

1. The contractor must provide the state, at a minimum, the following information for any Machine Learning, AI models, or Automated Decision-Making Technologies used to provide the system or services required in the contract:

6. A detailed description of the Machine Learning, AI model, or Automated Decision-Making Technologies to be used (e.g., model name, provider);

6. The specific purpose and scope of its use;

6. A description of the data inputs to be used and the outputs to be generated;

6. A summary of the contractor’s risk assessment, including how the contractor plans to mitigates risks of inaccuracy, bias, and data privacy violations;

6. Process workflow;

6. Network diagrams;

6. Large Language Model (LLM) frameworks and related services;

6. LLM data sources during training and ongoing;

6. Information security procedures;

6. Governance related to the LLM and infrastructure;

6. Audit procedures; and

6. Any other additional documentation or technical specifications regarding the technology that the state may request to evaluate compliance with state policies or assess and mitigate risk.

1. The state reserves the right to audit the Machine Learning solution, AI solution, and Automated Decision-Making Technologies used within the system and/or service. This audit may evaluate data usage, compliance, and the ethics and accuracy of the system's calculations and outputs.

1. The contractor shall be solely and fully responsible for the factual accuracy, completeness, and quality of all deliverables, regardless of whether AI was used in the preparation of the deliverable.

1. The contractor must warrant that all content, data, analysis, and recommendations generated by an AI system have been independently reviewed, fact-checked, and verified for accuracy and appropriateness by contractor qualified human experts prior to its submission of the content, data, analysis, and recommendations to the state.

1. The contractor warrants that all deliverables, data, and outputs provided under this contract shall be accurate, complete, and compliant, regardless of whether AI or Generative AI technologies were used to produce them. The contractor acknowledges that third-party AI provider disclaimers shall not shield the contractor from liability. The contractor shall indemnify, defend, and hold harmless the state against any third-party claims, losses, or damages resulting from AI-generated errors, omissions, systemic defects, or "hallucinations," and shall bear all costs associated with remediating such errors.

1. Data Isolation: The contractor must ensure that any approved use of an AI system is conducted in a secure, isolated environment (e.g., an enterprise-level instance) where "opt-out" from model training is enabled and enforced, and where no State Data is retained, logged, or used by the AI provider for any purpose other than to fulfil contractor’s obligations under this contract.

Web-Based Platform Requirements:

The contractor must provide a web-based platform that is available for Missouri Job Centers, Office of Workforce Development to use on location or virtually.

The contractor’s platform must be compatible with the operating systems of Windows 8.1+, MAC OS 10.15+, and Chrome OS 93+.

The contractor’s platform must be available for continued use in the event of an update or release made to the required operating systems.

The contractor’s platform must be compatible with the web browsers of Firefox 90+, Chrome 93+, Microsoft Edge 149.0.4022.98, or Safari 18.

The contractor’s platform must be available for continued use in the event of an update or release made to the required web browsers.

The contractor’s platform must have a responsive design compatible with desktop screens, tablets, and mobile devices.

The contractor’s platform must have access levels for administrators, proctors, and individual participants. All access levels must have login credentials.

The contractor’s platform must have a login troubleshooting option available for all users.

The contractor should provide an unlimited number of login credentials available for platform users.

The contractor’s platform must allow multiple users to log into the platform at the same time, including administrators, moderators, and individual participants.

The contractor’s platform must be compliant with ADA website requirements.

The contractor’s platform must have a service availability of no less than 99%.

The contractor must be responsible for all hosting requirements of the platform.

The contractor’s platform must assign a unique identity number or code to each participant.

The contractor’s platform must save earned certificates as a PDF.

The contractor’s platform must print or email certificates once complete.

The contractor’s platform must provide the ability for administrator level users to access, download, and print certificates for individual participants.

Technical Support Requirements:

The contractor must provide technical support for all platform users, including proctors and administrators, via a customer service hotline that is available by phone or email with a response time of less than twenty-four (24) hours.

The contractor must provide a toll-free telephone number for support.

The contractor must provide support Monday through Friday, 7:00 a.m. CST through 7:00 p.m. CST

The contractor’s technical support service hotline staff must be familiar with the various aspects of the program (assessments, curriculum, certification, and reporting).

The contractor must provide onboarding training for all administrator users and proctor users prior to assessment dates.

The contractor must offer training at local, on-site Missouri Job Centers, Office of Workforce Development centers interested in increasing their work readiness through citizen assessments and employer engagement.

The contractor must provide technical support immediately while participant testing is in progress to avoid delays or incomplete results.

Curriculum with Online Access Requirements:

The contractor shall provide a comprehensive online curriculum covering all topics included in the online assessment platform. The curriculum should include a quick assessment tool for participants to determine if further curriculum is required or if the participant is qualified for testing.

The contractor’s online curriculum shall have various levels of difficulty for the learning activities to be assigned to the individual participant.

The contractor’s online curriculum shall include practice sessions for individual participants to take an unlimited number of times prior to the assessment. The practice sessions must provide results at the end of the session for participants to review.

The contractor’s online curriculum system shall track individual participant’s time spent per session(s) and provide reports reflecting those details.

The contractor’s staff shall be able to assign individual participants to various levels of curriculum and track progress.

The contractor’s curriculum should track soft and hard skills and differentiate between soft and hard skills.

Online and Paper Assessment Leading to Certificates Requirements:

The contractor shall provide online and paper assessments covering topics of applied math, graphic literacy and workplace documents, in addition to specialized assessments focusing on applied technology, workplace observation and soft skills.

The contractor must provide customizable reporting tools that display customer results, certificate levels, demographic data and county in standard formats such as Excel.

The contractor must provide assessment scores that must lead to a certificate level that is available to customers online.

The contractor’s assessments should be available in Spanish for those requiring alternate language accommodations.

The contractor’s assessments shall meet ADA standards and provide participants with the option to increase font size, verbalize assessment questions, and allocate extra time.

The contractor should provide earned participant assessment certificates via PDF to participants within one (1) week of test completion.

Employer Engagement Requirements:

The contractor shall conduct employer outreach to ensure employers find value in certificates as meeting the skill requirements needed for their industries.

The contractor’s assessment results should align with the measure of skill determined by employers required for employees to succeed in workplace.

The contractor should provide education to employers to ensure employers understand how the certificate meets their job skill requirements.

System Maintenance and Enhancements Requirements:

The contractor must provide system maintenance of the career readiness, assessments and certifications web-based platform.

The contractor must provide system maintenance of the platform on weekends to ensure limited disruption to participants.

The contractor must update assessment and curriculum questions to remain current and relevant to workplace standards.

The contractor must notify the state agency and other entity users when enhancements or changes are made to the system that will affect normal operation. Notification must be sent via email at least one (1) week prior to system maintenance.

State Data:

State Data Location: All data associated with the service shall only be hosted, stored, and accessed within the United States. The contractor shall not store, locate, transfer, transmit or permit access to any portion of the state agency’s system instance, configuration, customization, documentation, or code related to the contractor’s services, nor any state agency data at rest, outside of the United States. The contractor shall also prohibit all personnel from storing state agency data on removeable media, including but not limited to USB drives, optical discs, or on mobile or personally owned devices such as tablets, mobile phones, or laptops. Storage of such data is permitted only on devices used exclusively within secured and controlled data centers.

Ownership: The state’s electronic data or information submitted by the state or its authorized users with authorized access to the solution/service shall be considered “State Data” for ownership purposes of the contract, as the state serves as either the owner or curator of the data submitted in the solution/service. “State Data” shall include: (a) the state’s data collected, used, processed, stored, or generated as the result of the Services; (b) personally identifiable information (“PII”) collected, used, processed, stored, or generated as the result of the services, including, without limitation, any information that identifies an individual, such as an individual’s social security number or other government-issued identification number, date of birth, address, telephone number, biometric data, mother’s maiden name, email address, credit card information, or an individual’s name in combination with any other of the elements listed herein; (c) protected health information (“PHI”) as that term is defined under the Privacy Rule, 45 Code of Federal Regulations (CFR) §160.103: and (d) all aforementioned data in (a), (b), and (c) without regard to whether such data is encrypted and/or de-identified. State data is and shall remain the sole and exclusive property of the state and all right, title, and interest in the same is reserved by the state.

“State Data” stored in the software and services being provided as a result of the contract shall also be known and treated by the contractor as confidential information in accordance with the contract, applicable contractor’s documentation, and applicable law.

Contractor Use of State Data: The contractor shall: (a) keep and maintain state data in strict confidence, using such degree of care as is appropriate and consistent with its obligations as further described in the contract and applicable law to avoid unauthorized access, use, disclosure, or loss; (b) use and disclose state data solely and exclusively for the purpose of providing the services, such use and disclosure being in accordance with the contract, any applicable Statement of Work, and applicable law; (c) not use, sell, rent, transfer, distribute, or otherwise disclose or make available state data for contractor’s own purposes or for the benefit of anyone other than the state without the state’s prior written consent; and, (d) not provide state data to any third party without the express written permission of the state. “…for the purpose of providing the services” as referenced above includes the right of the contractor to use state data to prevent or address service or technical problems, and verify service improvements, in accordance with the contract and the solution’s documentation, or in accordance with the state’s instructions.

End of Contract Extraction of State Data: At the end of the contract, the state shall have the right to extract its data from the solution for an additional sixty (60) calendar days from the date of contract expiration. The end of contract extraction of state data must be at no additional cost, conditions, or contingencies regarding the state’s right to extract the data notwithstanding the terms outlined in the contractor’s services agreement, and in the format that was mutually agreed to by the parties. If a term greater than 60 calendar days is required by the state to complete the state data extraction, the contractor agrees to work with the state to define the additional term and fees for the additional time.

Extraction of State Data Process Validation: The contractor must allow the state to perform routine validation of the extraction of the state’s data. The routine validation occurs to ensure the state's data will accurately transfer as prescribed to the state's designated extraction location and accommodates the volume of data transferred. The contractor shall assist the state by providing documentation and tools to allow the state to accurately process the extractions of the state’s data for the state’s continued use.

Backup and Recovery of State Data: Unless otherwise described in the contract, as a part of the services, the contractor shall maintain a backup of state data and perform an orderly and timely recovery of such data in the event that the production services may be interrupted in accordance with the disaster recovery and business continuity plans.

Return of State-owned Data: The contractor shall return state-owned data to the State of Missouri in a format mutually agreed to by the state and contractor and with a key or crosswalk to the information, where applicable, at no additional cost to the state. The state shall have the right to test the data extract provided by the contractor at a frequency determined by the state.

Software Piracy Prohibition: No state or other public funds payable under the contract shall be used for the acquisition,…