Procurement intelligence

QuoteSense

A quotation-analysis workflow that deliberately separates probabilistic document extraction from deterministic business logic. LLM output is normalized into typed structures before scoring, risk checks, anomaly detection, and comparative reporting.

Try live demo →View GitHub ↗

Media

Demo video & screenshots

Reserved space for a short product walkthrough and selected screenshots. Add the assets later without changing the page structure.

Problem

Procurement documents contain inconsistent structure, missing values, and commercial edge cases that are risky to evaluate with free-form LLM reasoning alone.

Solution

Extract structured fields, validate with schemas, then apply deterministic calculations and business rules before generating the final interpretation.

Architecture

Document ingestionLLM extractionPydantic validationDeterministic calculationsComparison + anomaly analysisHuman-readable result

Shared AI session

Connect the portfolio AI session

Set up the portfolio-level BYOK session once, then reuse it across the project pages.

No session

Temporary gateway session · reusable across portfolio projects

Open AI session →

← Back to projects