Experimenting With Gen AI at Savant
This post is the first in a series where we explore how we’re experimenting with, learning from, and incorporating Generative AI into all aspects of Savant. Our goal is to leverage AI to streamline workflows, enhance productivity, and create powerful automation tools that drive real business impact.
The Challenge: RFPs Are a Sales Bottleneck
Sales teams frequently encounter lengthy and time-consuming RFPs (Requests for Proposal), security questionnaires, and other documentation requirements that slow down deals. These documents often ask similar questions in slightly different ways, making the process highly repetitive. Completing them quickly and accurately is critical to closing deals on time, yet they remain a major pain point for sales and revenue operations teams.
Enter ChatGPT-powered Sales RFP Agents — an AI-driven solution that can automate the completion of these forms, reducing the time taken from days to minutes. In this post, we’ll walk through how you can build your own RFP agent to streamline your sales process and free up time for high-value work.
The Benefits of an AI-Powered Sales RFP Agent
Building an AI-driven RFP agent provides several key benefits:
- Massive Time Savings – What once took hours or days can now be done in seconds.
- Consistency and Accuracy – AI ensures that responses remain aligned with company-approved messaging.
- Scalability – Handle multiple RFPs simultaneously without burdening your sales team.
- Less Disruption – Sales reps can stay focused on closing deals rather than getting stuck in documentation tasks.
With a simple setup, you can build an agent that reads, understands, and completes RFPs using past responses as training data. Here’s how to do it.
Step-by-Step Guide to Building a ChatGPT Sales RFP Agent

1. Access ChatGPT’s Custom GPT Feature
ChatGPT now allows users to create custom GPTs tailored to specific tasks. To get started:
- Open ChatGPT and navigate to My GPTs.
- Click Create a new GPT.
2. Define the Agent’s Purpose
Give your agent a clear name and description. For example:
- Name: Sales Questionnaire Agent
- Description: “This agent automates the completion of sales RFPs and security questionnaires using knowledge from past responses. Upload a document, and the agent will generate answers for review.”
3. Provide Instructions and Training Data
Fine-tuning the agent’s behavior requires clear instructions. In the setup:
- Explain its role: “Complete sales questionnaires and RFPs based on past responses. Users can upload a document and receive a completed version.”
- Set expectations: “Always provide answers in a professional, concise manner. If uncertain, flag items for manual review.”
- Upload historical RFP responses to ground the AI in company-specific knowledge.
- 👀 Please note that you’ll need to upload more documents over time to improve the agent’s knowledge and keep it current.
4. Enable Key Capabilities
In the settings, enable:
- File Handling: Allow the AI to process uploaded RFP documents (Word, Excel, or PDF).
- Code Interpreter: Useful for reading and extracting structured data from files.
- Memory and Training: Store knowledge from previous responses to improve over time.
5. Deploy and Share the Agent
Once configured, you can choose to:
- Keep it private for your own use.
- Share it internally with your sales team via a private link.
- Make it public so anyone in your organization can access it.
How It Works in Practice

Using the Sales RFP Agent is simple:
- Upload an RFP document (PDF, Word, or Excel).
- Ask the agent to complete it.
- Review and refine the generated responses.
- Download the finalized document.
For example, we tested this with a cybersecurity questionnaire from a prospective customer. Within 30 seconds, the agent populated the form with accurate, on-brand responses. Previously, this process took two people two days — a 99% reduction in effort.

Future Enhancements: Continuous Learning
While the current setup requires manually uploading new RFP documents to enhance the agent’s knowledge, the next step is to integrate it with a shared drive like Google Drive. This would create a continuous learning loop, where the agent dynamically learns from completed RFPs and manual refinements without requiring manual uploads to the agent.
How This Would Work:
- A designated Google Drive folder stores all completed RFPs.
- ChatGPT periodically scans and ingests new documents to update its knowledge base.
- Manual refinements made by sales team members get automatically incorporated into future responses.
- Over time, the AI improves in accuracy, completeness, and adaptability, reducing the need for manual intervention.
This enhancement would make the agent smarter over time and further reduce human effort, ensuring that responses stay up to date with the latest company knowledge and messaging.
Final Thoughts: Transforming Sales Workflows With AI
This AI-powered solution is a game changer for sales teams, reducing friction in the sales cycle and ensuring faster deal closures. With just a 20-minute setup, you can save hours or even days of repetitive work.
If you’re constantly handling RFPs, security questionnaires, or compliance documentation, this is an essential tool. And if you have other repetitive tasks in your workflow, consider how custom AI agents can automate them, too.
Curious to see how Savant can further revolutionize your sales process with automation? Book a quick personalized demo, and we’ll show the true potential of analytics automation!