i2AO

AI reader for French public-tender documents

i2AO is a document-analysis SaaS for companies bidding on French public tenders (marchés publics). Instead of reading a tender's RC, CCAP, and CCTP PDFs clause by clause, a user uploads them and gets back a structured extraction of the deadlines, requirements, and obligations that matter. We built the extraction pipeline and the user and admin web apps around it.

  • 2024
  • France
  • PHP/MySQL
  • OpenAI Assistants API
  • PDF & DOCX ingestion
  • Plan-based subscriptions
Illustration of i2AO

The challenge

French public tenders come as a bundle of dense legal and technical PDFs: the Règlement de Consultation (RC) sets out the procedure and deadlines, the Cahier des Clauses Administratives Particulières (CCAP) covers contractual and financial terms, and the Cahier des Clauses Techniques Particulières (CCTP) lays out technical requirements. Reading all three by hand to find submission deadlines, eligibility conditions, and penalty clauses is slow and error-prone. i2AO answers this by routing each document type to an AI assistant with a checklist built for that document, so a bidder gets a structured summary back instead of a stack of PDFs to re-read.

Who uses it

Bidder
Uploads a tender's RC, CCAP, and/or CCTP files, runs an analysis, and downloads the extracted-clause report.
Admin
Manages subscription plans, user accounts, dashboard alerts, and tutorial content.

What we built

We built the document pipeline that uploads a bidder's RC, CCAP, and CCTP PDFs or DOCX files, sends each to an OpenAI assistant with a clause checklist tailored to that document type, and assembles the extracted findings into a downloadable PDF report, plus the user and admin web apps around it for accounts, plans, and history.

  • Multi-document analysis

    A single analysis run can cover any combination of RC, CCAP, and CCTP files uploaded together.

  • Document-type-aware extraction

    Each document type is sent to the AI with its own checklist: procedure and deadlines for RC, contractual and financial clauses for CCAP, technical requirements for CCTP.

  • PDF and DOCX ingestion

    Uploaded files are validated by MIME type and accepted as PDF or Word documents before analysis.

  • Generated report PDF

    Extracted findings are assembled into a new downloadable PDF rather than left as raw chat output.

  • Analysis history

    Past analyzes are kept per user, with the source files and the generated report available to revisit or delete.

  • Subscription plans

    Users sign up to Free, Gold, or Elite plans with usage limits, managed from the admin panel.

  • Dashboard alerts

    Admins can push a banner alert that users see on their dashboard.

  • In-app tutorial and FAQ

    Built-in tutorial and FAQ pages help users get started without outside support.

How it works

  1. User signs up and lands on the i2AO dashboard.
  2. User uploads the RC, CCAP, and/or CCTP files for a tender.
  3. Each file is sent to an OpenAI assistant with a prompt checklist built for that document type.
  4. The extracted findings for all uploaded documents are merged into a single report PDF.
  5. The user downloads the report and finds it again later in their analysis history.
Close-up of i2AO's uploaded-documents screen showing a tender's RC, CCAP, and CCTP files queued for analysis.
Close-up of i2AO's uploaded-documents screen showing a tender's RC, CCAP, and CCTP files queued for analysis.

Where the AI fits

Each uploaded RC, CCAP, or CCTP file is uploaded to the OpenAI Assistants API, which creates a document-specific assistant with a retrieval tool over that file. The assistant is given a checklist prompt tailored to the document type — procedure, deadlines, and submission rules for an RC; contractual, financial, and compliance clauses for a CCAP; technical specifications and quality-control requirements for a CCTP — and its extracted answer is pulled from the completed run once processing finishes. Results from all uploaded documents in one analysis are combined into a single generated PDF report.

  • OpenAI Assistants API: per-file upload, assistant, thread, and run
  • Separate extraction checklist per document type (RC, CCAP, CCTP)
  • Admin-editable prompt text per document type, falling back to a built-in default
  • PDF text parsing on the upload side via a PHP PDF parser library

Every part of the product

  • User web app

    PHP/MySQL

    Where bidders upload tender documents, run analyzes, manage their plan, and review history.

  • Admin web panel

    PHP/MySQL

    Manages plans, users, dashboard alerts, tutorials, and payment records.

  • Marketing website

    PHP

    Public site explaining the RC/CCAP/CCTP extraction services and pricing plans, in French and English.

How it’s built

Each tender document type gets its own OpenAI assistant and prompt rather than one generic summarizer, because an RC, a CCAP, and a CCTP need different checklists: procedural deadlines, contractual clauses, and technical requirements are different questions with different answer shapes. Admins can override each document type's prompt from the admin panel, with a built-in default as a fallback, so the extraction checklist can be tuned without a code change. Results from all documents uploaded in one run are merged and written back out as a new PDF, so a bidder gets one report instead of having to read the assistant's raw output per file. Analyzes, their source files, and the generated report are kept in a history table scoped to the user, and access to every endpoint runs through a bearer-token check tied to the user's account.

  • PHP
  • MySQL
  • OpenAI Assistants API
  • PDF parsing (Smalot PdfParser)
  • jQuery
  • Bootstrap admin theme

Related services: AI MVP development LLM integration

What clients say about working with us

Verified Fiverr reviews from our clients, including clients in France, quoted word for word.

5.0 98 verified reviews · Fiverr Top Rated
I enjoyed working with Venzpire on my project. The technical response meets my requirements and the ongoing communication is a strong point. I recommend this professional.
tristazFrance · Client on this project
I am very happy with everything Venkey did for me. I will certainly work with him in the future
alexsimion2019Repeat buyerFrance · AI Websites & Software
Absolutely top-notch experience! Venkey delivered a fully scalable AI Saas product with clean architecture and zero issues. Communication was fast, the work was professional, and the final product exceeded expectations. If you need serious backend work done right, this is your guy. Highly recommend
dnk3201Repeat buyerIndia · AI Websites & Software
The team understood the requirement and was able to deliver before time. They also took care of the deployment. Great job !
debabratatha136India · AI Websites & Software

Read all 98 client reviews

Questions about i2AO

What does i2AO do?

i2AO is a document-analysis SaaS Venzpire built for a client in France that reads the RC, CCAP, and CCTP documents from a public tender and extracts the clauses, deadlines, and requirements a bidder needs, instead of leaving them to read dense legal PDFs by hand.

How does i2AO read tender documents?

Venzpire built i2AO so that each uploaded document is sent to an OpenAI assistant with a checklist prompt specific to its type: procedure and deadlines for an RC, contractual and financial clauses for a CCAP, and technical requirements for a CCTP. The extracted findings are then merged into a single downloadable PDF report.

Who uses i2AO's admin panel?

Venzpire built a separate admin web panel for i2AO where staff manage subscription plans, user accounts, dashboard alerts, and the per-document-type extraction prompts, keeping that configuration apart from the bidder-facing app.

Your product could be next.

Tell us what you are building. Venktesh reads every application himself and replies within one business day.

Apply for a strategy call