TenderCheck AI is an intelligent SaaS platform purpose-built for SMEs bidding on public contracts. Winning a government tender requires meticulously matching every requirement clause in the official tender document (known as a pliego in Spanish public procurement). Manually reviewing dozens of pages of dense legal and technical language is error-prone and expensive. TenderCheck AI automates that process end-to-end: upload a tender PDF, let the AI extract every mandatory and optional requirement, then upload your proposal to receive a structured compliance validation report with clickable source citations — all in minutes.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/elecodes/TenderCheck-AI/llms.txt
Use this file to discover all available pages before exploring further.
Key Features
AI Tender Analysis
Powered by Google Gemini 2.5 Flash via Google Genkit, the platform adopts a Legal Auditor persona to extract technical clauses from tender documents, distinguishing between OBLIGATORIO (mandatory) and OPCIONAL requirements with high precision.
Proposal Validation
A Senior Evaluator AI persona cross-references your vendor proposal against every extracted requirement, understands technical synonyms and partial compliance, and delivers a structured pass/fail report with detailed reasoning per item.
Clickable Citations
Every extracted requirement carries a source page badge. Click it to open a modal displaying the full page text with the AI-identified fragment highlighted — eliminating blind trust in AI output.
Large PDF Support
Page-based chunking handles documents up to 80+ pages. PDFs exceeding 15 pages are automatically split into 10-page chunks, processed in parallel (up to 3 concurrent), and aggregated into a unified result.
Industry Validation
Dynamic scope filtering adapts validation to different sectors (Digital Services, Construction, and more) using database-driven keyword presets, so irrelevant clauses are flagged and in-scope requirements are prioritised.
Secure Auth
Hybrid authentication strategy combining HttpOnly Cookies (XSS protection) with a Bearer Token fallback, plus Google OAuth with PKCE (Authorization Code flow — no access token in the URL) and JWT-signed sessions.
Persistent History
All analyses and validation results are stored in Turso (distributed LibSQL), giving you a searchable, deletable history of every tender and proposal you have evaluated.
PDF Export
Generate branded, professional PDF reports of any analysis or validation result for offline review, team sharing, or regulatory record-keeping.
How It Works
Upload the Tender Document
Upload the official tender PDF (pliego) through the dashboard. The backend accepts files up to 50 MB and extracts the full text using a page-aware PDF parser.
AI Requirement Extraction (Legal Auditor)
Gemini 2.5 Flash adopts the Legal Auditor persona and scans the extracted text for obligation signals (“deberá”, “obligatorio”, “se requiere”). Each identified requirement is classified as mandatory or optional, assigned a source page, and stored as a vector embedding for semantic search.
Upload Your Proposal
With the tender analysis complete, upload your vendor proposal PDF. The platform retrieves the relevant requirements via vector semantic search (
gemini-embedding-001, 3072 dimensions) to focus validation on what matters.Compliance Validation (Senior Evaluator)
Gemini 2.5 Flash re-engages as the Senior Evaluator, comparing each requirement against the proposal text. The result is a detailed compliance report: per-requirement status, confidence scores, justifications, and an overall pass/fail summary — ready for export.
Tech Stack
| Layer | Technology |
|---|---|
| Frontend | React 19 + Vite + Tailwind CSS |
| Backend | Express 5 + TypeScript (Clean Architecture) |
| Database | Turso / LibSQL (distributed edge SQLite) |
| AI | Google Genkit 1.28 + Gemini 2.5 Flash |
| Embeddings | gemini-embedding-001 — 3072-dimensional vectors |
| Auth | JWT + HttpOnly Cookies + Google OAuth PKCE |
| Observability | LangSmith (tracing, prompt versioning, performance) |
| Deployment | Vercel (frontend) + Turso (database) |
Live Demo
A fully deployed instance is available at https://tendercheckai.elecodes.online. You can sign up for a free account and test the complete analysis and validation flow using the sample documents provided indocs/Testing_docs/ of the repository:
Pliego_Tender_IT_Security.pdf— use this as the tender document inputOferta_Offer_IT_Security.pdf— use this as the proposal document input
TenderCheck AI ships with 92%+ global test coverage and 100% domain layer coverage, enforced as a mandatory CI quality gate. Every core business rule, entity, and use-case is fully covered by Vitest unit tests.
Next Steps
Quickstart
Get a local development instance running in under 10 minutes with full AI analysis capability.
API Overview
Explore the REST API endpoints for tender analysis, proposal validation, authentication, and history management.