RFI_ AI Contact Center Solutions Dec 2025 (1).docx

Request for Information (RFI): AI Contact Center Solutions

The Government is seeking information from industry on AI-enabled solutions to support government contact centers, including virtual voice agents, chatbots, and other advanced technologies (e.g., generative AI, agentic AI).

The purpose of this RFI is to:

1. Understand the current market capabilities, emerging technologies, and approaches to delivering fast, accurate, and autonomous customer service. The Government’s objective is to identify AI capabilities that autonomously, or to the greatest extent possible, automate contact center interactions.

2. Obtain feedback on an initial, incomplete draft of the Performance Work Statement (PWS). This version is provided for review and comment only; it does not represent a finalized requirement. Industry insight is sought to help shape, clarify, and strengthen the PWS prior to completion.

Responses will inform a future solicitation for AI solutions to handle multi-agency contact center inquiries. This RFI is for information and planning purposes only; it is not a solicitation or request for proposal.

Instructions to Respondents

· Provide clear, concise, and complete responses to each question. Responses are limited to 200 words per question

· Provide responses to Kenny Yiu at kenny.yiu@gsa.gov and Kelsie Kharibian at kelsie.kharibian@gsa.gov Company Information

· Legal company name, address, and primary point of contact.

· Years in business and experience delivering AI contact center solutions for government or large commercial clients.

· Size of the company and number of employees supporting AI or contact center solutions.

Capabilities & Functionality

1. Please describe your AI contact center solution and how it resolves or supports multi-channel customer interactions (voice, chat, email, and digital channels). Include a description of its architecture, automation capabilities, and approach to achieving end-to-end resolution for Tier 0 and Tier 1 interactions (as defined in the PWS).

2. Explain how your solution can integrate with government systems, third-party platforms (e.g., Salesforce, ServiceNow, CCaaS), and other AI solutions. Include your approach to APIs or other interfaces, and any experience layering your AI solution on top of existing systems or even other AI solutions.

3. Describe how your solution ingests, organizes, and retrieves knowledge from multiple sources (FAQs, documents, websites, knowledge articles, etc.), and how it ensures responses are accurate, grounded, and contextually relevant to the user’s request.

4. Explain how your solution leverages state-of-the-art or emerging AI technologies (e.g., agentic AI, generative LLMs, autonomous orchestration frameworks) to improve automation, accuracy, and customer experience. For each technology, describe the specific functions it performs, the use cases it addresses, and the operational outcomes achieved.

5. Is the solution FedRamp authorized, and at what impact level (Low, Moderate, or High)?

6. How does your solution measure and maintain AI accuracy and quality over time?

7. How does your solution prevent AI outputs from generating unsafe, biased, or inappropriate content?

8. How does your solution manage human-in-the-loop interventions or escalations?

9. What criteria or confidence thresholds trigger escalation to a human agent? How customizable are these thresholds and escalation workflows?

10. What is the maximum expected concurrent user capacity your system can handle for both voice and chat?

11. How does your solution handle peak demand or rapid scaling requirements?

12. What monitoring, auditing, and reporting capabilities are included for QA purposes?

13. Please describe the flexibility and configurability of your AI solution, including how it can be adapted to different workflows, business rules, languages, channels, or evolving requirements with minimal technical effort.

14. When your AI fails to resolve an issue, how does it detect failure, recover gracefully, and avoid customer frustration or infinite loops?

Innovation

15. Highlight how your solution compares to the current AI industry.

16. What future capabilities or innovations are planned for the next 2 years?

Metrics

17. Based on the requirements described in the draft PWS and your solution’s capabilities, please estimate the percentage of total Tier 0 and Tier 1 inquiries that can be resolved end-to-end by the AI without human intervention. Include any assumptions or dependencies that may affect these estimates. For the purposes of this question, ‘resolved' means the AI successfully handled the inquiry, as confirmed by the customer, without human intervention or escalation.

18. Please provide metrics or results from prior deployments (e.g., resolution rate, average handle time reduction, user satisfaction)? Please clearly define the specific metric used, as metrics may vary across AI systems.

Cost and Licensing

19. What pricing models do you offer (per-agent, per-conversation, subscription, usage-based, enterprise license)?

20. Please provide the specific pricing for each model offered, including any volume-based discounts, tiers, or additional fees.

21. Are there additional costs for integration, customization, or ongoing AI training? If so, please provide the specific pricing for these costs.

PWS Feedback and Recommendations

22. Please provide feedback or suggestions on how the draft PWS, including its technical solution requirements and performance metrics, could be improved to better achieve the government’s stated objectives.