Currently exploring new opportunities · Paris

Alexandre Mourin

Strategy Consulting ·
Builds AI Systems

Advising enterprise clients on strategic and innovation missions at Tata Consultancy Services. Building AI systems and agents on my own time.

Builds →Experience Get in touch

About

Strategy by training. Builder by habit.

From Montpellier to Sydney, from Paris to Warsaw — strategy at the day job, AI systems on every evening.

Co-founded an AI consulting agency in 2022. Spent 9 months solo in Sydney. Six months at Renault Warsaw on corporate strategy. Now advising enterprise clients on strategy and innovation at Tata Consultancy Services.

My background is strategy — scoping, competitive analysis, board-level recommendations. But I'm convinced AI is reshaping how every industry operates. That's why I've built AI systems on my own time — multi-agent pipelines, LLM orchestration, a trading bot, a 250-member community. Not to put them on a CV. To understand how these technologies actually work.

Based in

Paris 16e

Open to

Anywhere

Education

MBS — Master + MSc Consulting & Digital Transformation

Languages

French (native) · English (C2)

Experience

Strategy at the core. Systems thinking at every step.

Four roles, three countries — consulting, performance, retail strategy, network development.

Tata Consultancy Services

— Strategy & Innovation Consultant

Aerospace · Automotive · Pharma

Mar. 2026 – PresentParis, France

Scoped and ran strategy consulting engagements for enterprise clients (aerospace, automotive, pharma) — from diagnosis to recommendations presented to executive committees.
Assessed AI maturity and identified transformation opportunities, translated into decision-ready roadmaps for C-level stakeholders.
Produced executive deliverables (strategic reports, board / C-level decks) used as the basis for investment and innovation decisions.
Conducted competitive and sector analyses framing clients' positioning and strategic trade-offs.

Impulse Analytics

— Performance & Growth Analyst

PULLIN · GAIA · Decathlon · €150k+ annual media spend

Jul. 2025 – Jan. 2026Paris, France

Built automated workflows for Google Ads monitoring and reporting — cut manual reporting time by ~60%.
Managed and optimized client portfolios representing +€150k in annual media spend — 20–30% average ROI improvement.
Deployed data-driven acquisition strategies on PULLIN, GAIA, Decathlon.

Repair&Co

— Strategy & Transformation (independent mandate)

Multi-site retail · National expansion

Sep. 2024 – May 2025Montpellier, France

Contributed to redefining the strategic model: internal diagnosis, competitive analysis, repositioning recommendations.
Work integrated into an expansion plan leading to 4 new outlets in 18 months, ~30% capacity increase.

Renault

— Business Analyst — Network Development

Renault Group · Mobilize

Apr. – Sep. 2024Warsaw, Poland

Mapped and segmented ~100 Renault dealerships across Poland: opportunity scoring, development recommendations.
Designed launch strategy and deployment plan for the MOBILIZE app on the Polish market.
Network optimization recommendations integrated into Renault Group's 2025 strategic plan.

Paris · 2026 → now

Tata Consultancy Services

Warsaw · 2024

Renault Group

Builds

I learn by building.

Every new technology that ships, I want to understand it — not by reading about it, but by making something with it. Systems I build, use and operate every day.

Open-source on github.com/alexandremourin — public reference architectures credited in each repo where applicable.

01live

Agent Pipeline

Multi-agent orchestration system with MCPs

Four-agent architecture: Scout sources entries across 12+ channels, Scorer evaluates on 11 weighted criteria, Adapter generates tailored outputs, Reviewer runs human-in-the-loop QA.
Full state tracked in Notion. 300+ entries processed automatically.

Claude CodeMCPNotionPython

02live

Ponder

Self-correcting agentic RAG

Routes queries, grades retrieval relevance and rewrites weak queries before answering.
Answers with inline citations, then runs a self-critique pass to catch its own errors.
0.94 retrieval recall on the evaluation set.

PythonLangGraphLocal embeddingsChroma

View on GitHub

03live

needle

Semantic code-search MCP server

MCP server that returns the right code snippets (file:line) instead of whole files.
Hybrid retrieval — BM25 + dense + RRF — over tree-sitter chunks, fully local.
~15–44× fewer tokens at ≥0.9 recall.

PythonMCPtree-sitterBM25 + dense + RRF

View on GitHub

04live

maestro

Hierarchical multi-agent research

Planner decomposes the task, parallel workers run web searches, a synthesizer merges findings.
Produces a sourced answer with citations from the gathered evidence.

PythonasyncioWeb search

View on GitHub

05live

patch

End-to-end bug-fixing SWE agent

Locates the bug, edits the code, runs the tests and iterates until they pass.
Outputs a clean git diff through an agent–computer interface in an isolated workspace.

PythonAgent–computer interfaceIsolated workspace

View on GitHub

06live

LLM Council

Multi-LLM deliberation layer

Built on an open-source deliberation framework — implemented, customized and integrated into daily workflow.
Connects 4 LLMs (Claude, GPT, Gemini, Groq) in a 3-stage flow: parallel generation, anonymized peer-ranking, chairman synthesis.

PythonFastAPIOpenRouterClaude APIReact

07live

Jarvis

Personal AI Chief of Staff

Orchestrates parallel projects with calendar awareness.
Daily briefing (/sitrep) and deep-work mode (/labmode) for task triage and strategic review.
File-based state per project, permissions matrix enforcing read/write boundaries.

Claude CodeMCPGoogle Calendar APIPython

08live

Flywheel

Automated content pipeline — Instagram

End-to-end content system handling scripting, generation, scheduling and publishing.
Runs daily in production with minimal manual input.
Consistent cadence across multiple formats — no missed publish since launch.

PythonClaude APINotion APIInstagram Graph API

09in validation

Trading Bot

Paper trading · walk-forward validation

ML long / short signals with LLM confirmation and automated risk management (kill-switch).
Runs autonomously in paper mode on crypto perpetuals — no live capital.

PythonFastAPIGroqBinance API

Stack