Leobit
AI-Powered Solution that Automates RFP Workflows
Custom development of an intelligent sales assistance tool for a Caribbean-based IT company
ABOUT the project
- Client:
- Caribbean-based IT services company
- Location:
-
The Bahamas
- Company Size:
- 50+ Employees
- Industry:
-
IT Services and Consulting
- Solution:
- Custom AI-Powered Software
Leobit planned the architecture and the functionality on an AI-powered request for proposal (RFP) system for a Caribbean-based IT service company. Our specialists provided the customer with an architecture proposal and proceeded with active development. The solution will help our customer find potential partners across government and procurement portals, assess and categorize promising cooperation possibilities, and generate an RFP draft.
Leobit already had experience in building an AI sales assistant, which was very important for this project. They came in with a clear sense of what to build and how, which cut the planning phase significantly.
Customer
The customer is a Caribbean-based IT services and consulting firm operating across multiple regional jurisdictions, including Bermuda, the Cayman Islands, and the Bahamas. The company provides technology services to government and enterprise organizations across the region.
Business Challenge
The customer identified request for proposal (RFP) opportunities manually across multiple government and procurement portals in the Caribbean region. The process was time-consuming, inconsistent, and made it difficult to prioritize the most relevant opportunities efficiently.
They needed a solution that would replace manual portal monitoring with an automated, AI-driven pipeline.
Why Leobit
Leobit brought direct experience in building AI-driven sales and procurement tools, including its internal RFP scoring and proposal generation system, Leona. This expertise matched the customer’s preference for a lightweight, low-maintenance solution built on Microsoft’s AI ecosystem.
Project
in detail
The project is ongoing. We outlined the solution’s core functionality and architecture while working on a project proposal and proceeded with active development.
The Leobit team analyzed the customer’s requirements to come up with a vision of a system that would meet their strategic goals. We outlined an architecture of a solution built on Microsoft Azure’s ecosystem, using serverless and AI-native servers. We also developed scrapers for multiple portals for searching customers across the Caribbean region. Finally, our team created a standardized RFP structure in the form of a template.
We proposed a solution based on several services from the Microsoft stack. The current blueprint includes the following services:
- Azure Functions covers serverless compute for daily scraping jobs, processing pipelines, and digest generation.
- Azure OpenAI provides models for AI-powered summarization, classification, scoring, and draft response generation.
- Azure Storage is used to store documents and attachments.
- Custom CI/CD pipelines cover automated deployment and versioning.
We are currently in the active development phase of the solution, building directly on the architecture blueprint we have established. Our specialists implement custom keyword logic behind the solution’s scrapers that automatically identify relevant opportunities across different platforms. We also configure the model to score the relevance of cooperation opportunities, provide their summaries, classify them by categories and priority, and automatically generate tailored requests for proposals. In the future, we will integrate the system with email services, which will allow the customer to send daily digests to configured recipients.
AI-Powered RFP Scoring
Automated scrapers monitor multiple RFP portals daily, collecting new opportunities and extracting metadata including deadlines, issuing organizations, submission methods, and relevant attachments. After that, an AI model analyzes these opportunities, categorizes them based on relevance, category, and priority score. At the end of the day, a sales team receives a digest with the outcomes of the solution’s research. The AI model also generates an RFP draft for high-priority prospects, which can be immediately sent to them upon approval from the sales team. This saves the time of the customer’s sales agents while helping them quickly identify and categorize the most relevant cooperation opportunities.
Azure-Native AI Pipeline Architecture
The solution runs entirely on Microsoft Azure’s serverless and AI-native services. Azure Functions handles daily jobs in response to particular triggers, while Azure OpenAI handles the system’s core logic for evaluating and responding to RFP opportunities. All the data is stored and organized in the Azure Storage. Such architecture removes the need for dedicated infrastructure management and scales on demand with the volume of incoming RFPs.
Unified Schema and Match Reasoning
To make every RFP comparable, the system maps collected source data into a unified schema that captures title, deadlines, issuing organization, submission method, attachments, and category. This standardization is what enables consistent relevance scoring across very different portal formats.
On top of the schema, Azure OpenAI generates a short, human-readable match reasoning paragraph for each RFP. This shortens the review cycle and turns the AI’s decision into an auditable recommendation.
Technology Solutions
- Custom configurations for Azure OpenAI models enabling them to analyze, categorize, and score RFP opportunities, as well as generate RFP drafts for high-value prospects.
- Unified schema for evaluating and organizing RFP opportunities according to pre-configured settings.
- Flexible and cost-effective serverless architecture based on Azure Functions.
- Custom CI/CD pipelines ensuring automated software versioning, delivery of new features, and ongoing development of the system.
Value Delivered
- Ongoing work on a comprehensive and highly-customizable solution that will address the main challenges of the client’s sales team.
- Sales workflow automation through the use of customly configured AI capabilities.
- A 3x acceleration in the customer’s sales team efficiency.
- An innovative and structured approach to sourcing and categorizing RFP opportunities.
