Blockchain & Cie: Software & AI Engineering Studio
Software and AI agents that hold up in production.
Blockchain & Cie audits, stabilizes and builds your software systems and AI agents. Antoine Delamare leads every engagement, with 10 years in pharmacy and regulated traceability behind him.
2 CVEs fixed
On an industrial ERP in production (CVE-2023-30533, CVE-2024-29041).
~270 commits across 9 repos
On a traceability feasibility study funded by ESA through Parametry.
175 commits, 77% of the repo
On an AI agent stack in production, over about 5 months.
~103 pytest tests
Covering configuration contracts and agent flows in production.
Two ways to work together
Fix a system that already exists
Your code works, but nobody wants to touch it anymore. We audit it, stabilize it and hand it back to your team.
- An AI-generated repo you need to bring up to standard.
- Incidents or regressions that keep coming back.
- A team that needs structured AI development practices.
Build something new
You have a validated need. We design, build and deploy it, with bounded autonomy wherever an AI agent is involved.
- A business workflow to automate with an AI agent.
- A POC or MVP to ship in weeks, not months.
- A dApp project with smart contracts.
The 6 offers
3 offers to fix, 2 to build, 1 for continuity. Each offer states the expected outcome, deliverables and eligibility criteria, so you can tell quickly which one fits. The details live on each path page.
Repo Rescue: Sanity Check
AI-generated code audit and hardening
A prioritized risk report on your repo. You know what breaks, what's urgent and what's left to do before deciding.
1 to 2 days
Repo Rescue: Rescue Sprint
targeted codebase rescue with a locked scope
A stabilized, tested and documented codebase, ready for your team. No surprises: the scope is locked after diagnosis.
5 to 10 days
AI-Driven Dev Workflow Setup
setting up AI-assisted development practices
Your team adopts clear conventions for coding with AI tools: documented context files and integrations that hold up. Generated code stays consistent from one developer to the next.
3 to 5 days
AI Agent Deployment
deploying an AI agent in production, with bounded autonomy
An AI agent in production on your business workflow, with bounded autonomy and human control. Supervised, tested, documented. Your team stops repeating the workflow by hand.
2 to 4 weeks
POC / MVP / dApp Build
full build, from design to deployment
A working, deployed product, ready for an investor demo or a real user test.
POC: 4 to 6 weeks. MVP: 3 to 4 months.
AI Agent Retainer
ongoing maintenance and improvement of AI agents in production
Your AI agent stays stable and up to date, and gains new workflows every month, without a full-time hire.
monthly, 3-month minimum
Recent work
An industrial ERP brought back into production. An AI agent infrastructure at a European cloud provider. A traceability study funded by ESA. Details anonymized where required.
Industrial ERP, back in production
Industrial technical textile SME, internal ERP built with AI tools.
- 2 CVEs fixed (CVE-2023-30533, CVE-2024-29041)Prototype pollution in a file-generation library and path traversal in the web server, both publicly listed.
- 46 tests (Mocha + supertest)Automated test suite added with the stabilization release.
- 0 remaining external CDN dependenciesThe affected libraries are vendored locally; the application keeps working offline.
This proof documents an audit and a stabilization of existing code. It does not demonstrate a full architecture overhaul or a multi-environment deployment.
- Node.js
- Express
- Mocha
- supertest
European cloud provider, AI agent infrastructure
European cloud provider offering a sovereign infrastructure product.
- 175 commits, 77% of the repo over about 5 monthsAuthorship volume on the main repository of the production agent stack.
- ~103 pytest testsTest coverage on configuration contracts, end-to-end HTTP flows and a dedicated guardrail separating dialogue from business logic.
- 22 tools exposed to the agentNumber of runtime tools (messaging channels, filesystems, MCP) available to the agent within the deployed scope.
This proof documents an execution and hardening infrastructure for agents. It does not document the business outcomes achieved by the end clients of those agents.
- Docker
- Python
- FastAPI
- Node.js
- MCP
- Matrix E2EE
Cotton traceability, ESA feasibility study through Parametry
Feasibility study funded by the European Space Agency (ESA), run through Parametry: tracing organic cotton from field to fiber.
- ~270 commits across 9 GitLab reposContribution volume across every repository of the project, between October 2025 and August 2026.
- 105 tests on the smart contract suiteHardhat test coverage on the identity token, the transaction certificates and the dual-verification oracles.
- ~196 tests on the satellite pipelinePython test coverage on the satellite data processing chain and the compliance score computation.
This proof documents a feasibility study and an end-to-end proof of concept. It does not document a large-scale production deployment or an obtained regulatory certification.
- React
- Vite
- Solidity
- Hardhat
- FastAPI
- PostGIS
- Python
AI-assisted writing SaaS, stabilization
AI-assisted novel-writing SaaS, a React/TypeScript application in production.
- 86 commits, 20 test cyclesVolume of work across 4 weeks of beta testing and stabilization.
- 575 files stabilizedScope of the React/TypeScript application covered by the engagement.
- 14 bugs fixedIncluding an authentication race condition and cases of incomplete generation.
Secondary proof of qualification and stabilization work. It does not document the initial product design or its commercial results.
- React
- TypeScript
- Firebase
- Netlify Functions
How we work
Every engagement follows the same logic: scoping, diagnosis, work within a locked scope, handoff.
- 01
Scoping
A call to understand the system, the context and the expected outcome.
- 02
Diagnosis
Analysis of the code or the need, returned as risks and priorities.
- 03
Work within a locked scope
The scope is fixed after diagnosis. No silent drift once the engagement starts.
- 04
Handoff
Documentation, tests and a handover to your team at the end of the engagement.
Regulated environments
10 years in pharmacy, a regulated sector, before development. That background shapes the work on traceability and compliance.
- Doctor of pharmacy (PharmD), 7 years of pharmacy practice before the career change.
- A habit of quality control, documentation and regulatory accountability.
- Blockchain & Cie guarantees no certification or compliance outcome: the scope of each engagement states it explicitly.
Managing director and technical lead
Antoine Delamare
Antoine leads every engagement on the technical side. Before development, he spent 7 years in community pharmacy, a regulated sector. He works remotely in French, English and Spanish, and stays accountable for the outcome, even when a specialist covers a specific part.
- Doctor of pharmacy (PharmD)
- French, Spanish, English
- 100% remote
Frequently asked questions
How much does an engagement cost?
The price depends on the exact scope. Each offer states an indicative duration. Pricing follows the scoping call or the Sanity Check, and you approve it before any commitment.
What happens during the call?
You describe the system and the expected outcome. Antoine asks the technical questions. By the end of the call, you know which offer fits your situation, or whether none does.
Do you work alongside existing internal teams?
Yes. Most engagements run alongside your team or a product contact, with a documented handoff at the end.
Who actually works on my engagement?
Antoine Delamare leads every engagement on the technical side. When needed, specialists cover a specific part of the scope under his responsibility.
Can an AI agent act without supervision?
No. Deployed agents run with bounded autonomy and human control rules on sensitive decisions. The exact scope is defined with you.
Do you work outside France?
Yes, remotely, in French, English or Spanish. Blockchain & Cie is a French entity.
What happens after a Sanity Check if I don't continue?
The risk report belongs to you, with no obligation to continue.
A system to fix or to build?
Describe your context on a call. You leave with a clear recommendation: no commitment, no generic diagnosis.