# Alain Lorenz — CTO & co-founder — AI & MLOps

> Co-founder and CTO. I build complete products and run their AI models in production, from raw data to GPU inference.

- Between Geneva and Lyon
- Languages spoken fluently: French · German · English
- Web: https://ala.inlore.nz/en/
- LinkedIn: https://www.linkedin.com/in/alain-lorenz-4029862b/
- Instagram: https://www.instagram.com/wizopunker/
- CV (PDF): https://ala.inlore.nz/cv/alain-lorenz-cv-en.pdf
- ATS-friendly CV (PDF): https://ala.inlore.nz/cv/alain-lorenz-cv-ats-en.pdf
- Language: [Français](https://ala.inlore.nz/index.md), [Deutsch](https://ala.inlore.nz/de/index.md)

**Full stack, quite literally: from the phone down to the GPU.** I build complete products and run their AI models in production — from raw data to GPU inference. Co-founder and CTO of JobyPepper and Ozly.

## The CTO who still commits.

As co-founder and CTO, I have led teams of up to ten people without ever letting go of the keyboard: I am still the top contributor to the API I started in 2016. Since 2023, my main playground has been AI in production — training, deploying, serving and optimising models that answer real users. My favourite subjects: LLMs, RAG and fine-tuning. And the paperwork that keeps the company running — GDPR, DAC7, tax, innovation tax credit — goes through me too. All between Geneva and Lyon.

- 10 years on the same product
- 10,600 commits under my name
- 36 git repositories
- ÷4 GPU cost after optimisation

_Yes, a CTO who commits. The species is not extinct._

## The path of a request

I built every layer it crosses. Half of the trip happens on the AI side.

### The screen

`app.tap("Search")` — React Native / Expo mobile apps, including a complete business app, taken all the way to the stores: iOS and Android builds, submission, listing copy. React web interfaces for admins, clients and freelancers.

_It all starts in a pocket._

### The API

`POST /search` — Node.js / Express, business workflows, and a financial dispatch engine: a single client payment automatically becomes every freelancer’s invoice, the fund transfers and the payouts.

_Where the money flows. No pressure._

### The queue

`rabbitmq.publish(search)` — RabbitMQ between the API, workers, scheduler and real-time services. Each service does one thing, asynchronously.

_Everything gets through eventually. In order._

### The search (AI)

`vector.search(k=20) + filters` — Embeddings, vector search combined with structured filters (location, sector, contract), conversational RAG.

_Find the needle. Then explain why it is a needle._

### The model (AI)

`camembert.classify → llm.answer` — Fine-tuned CamemBERT classifiers, sourcing and matching agents, MCP servers that plug LLMs into real business data.

_The part everyone talks about._

### The GPU (AI)

`triton.infer(gpu:0)` — Inference served by Triton on GPU, CUDA containers.

_The part that gets expensive when nobody is watching._

## A model in a notebook is a promise. In production, it is a service.

MLOps is everything in between. I cover the whole cycle.

1. **Collect** — Scheduled pipelines that aggregate, clean and deduplicate data from multiple sources.
2. **Train** — Dataset building, fine-tuning classifiers by sector and contract type.
3. **Package** — Versioned models, CUDA Docker images, published to a container registry.
4. **Serve** — GPU inference, an embedding service on Kubernetes, vector search.
5. **Operate** — CI/CD, observability on every service, continuous optimisation: embeddings cut down to 512 dimensions, a memory leak hunted down and fixed.

_Step 5 never ends. That is expected._

## How I work

Projects run end to end with AI, and a human at every checkpoint.

### Pipeline of a project, from idea to production

1. The project is declared in Notion + Figma. (human)
2. It is reviewed and approved by a human. (human)
3. LLMs build it, plugged into the code and its context. (AI)
4. I verify, test and fix. (human)
5. Shipped to production: Docker images, Kubernetes deployment. (CI/CD)
6. Errors flow into Sentry, readable by the agents. (agents + human)

My real speciality is what happens after the release: monitoring, diagnosing, stabilising — in record time.

_The agents read Sentry before my first coffee. The fix often ships before it gets cold._

### Start from the need

- Before the first line of code: understand the business, who will use the tool, to do what, under which constraints.
- Derive the user journeys from that, then the UX/UI. The interface follows the need, not the other way round.
- An in-house design system so the whole product speaks one language.

### Leading

- Technical leadership of a team of up to ten people.
- Architecture decisions, delivery planning, product / tech trade-offs.
- Preparing innovation funding applications.
- Turning an operational need into a tool, and an AI experiment into a service that runs.

### Working with AI, not next to it

- Coding agents in my daily workflow, with versioned instructions in every repository.
- MCP servers built on three products to plug LLMs into business data.
- Production errors are readable by the agents: diagnosis and a proposed fix, right away.

### Running the company

- GDPR compliance and the filings that come with it.
- DAC7 exports and filings, the EU tax reporting duty of collaborative platforms.
- Corporate taxes and statutory filings.
- Tracking the projects eligible for the French innovation tax credit (CII).

_This site was designed with two AIs. I wrote the brief, settled their disagreements and proofread every sentence. That is called management._

## Projects

### [JobyPepper](https://www.jobypepper.com)

Technical co-founder · CTO · since 2016

The platform that connects companies with freelancers for short missions.

- Web, mobile, API and infrastructure: the whole product ecosystem.
- Two mobile apps published on the iOS and Android stores, from build to listing: versions, submission, compliance fixes, store copy.
- Payment dispatch, not just a “pay” button: the client pays once, the platform generates every freelancer’s invoice, splits the funds and triggers each payout. Automatically.
- AI agent for candidate sourcing and matching, connected through MCP.

### [Ozly](https://hello.ozly.ai)

Technical co-founder · CTO · 2023 · on hold

AI applied to job search.

- Full pipeline: multi-source ingestion, CamemBERT classification, vector index.
- Semantic search and conversational RAG, in an embeddable widget.
- Optimised GPU inference: cost divided by four.

### On the side

#### p.oint

Client project · 2026

Attendance check-in for schools and youth centres: tap your name on the tablet, “I’m here”, “I’m leaving”.

- Tablet kiosk as a PWA that keeps checking people in offline and syncs when the connection returns.
- Staff dashboard: sessions, statistics, Excel-friendly CSV import/export, tablets paired by QR code.
- Delivered turnkey: Node + SQLite API, Debian install, HTTPS and daily backups.

#### Origin AI

Personal project · 2026

Videos of real workplaces turned into training data for robotics.

- Chunked video upload, ffmpeg processing, LLM tagging.
- Automatic face blurring on GPU, dataset creation pipeline.

#### EV Comparator

Side project · 2026

After how many kilometres does electric become cheaper?

- Daily open data ingestion, cost-per-kilometre model.
- Resale value shown as an estimate, never as the truth.

## Off-piste.

I am not just a geek. Above all, I am someone who has fun — on the water, on the snow, on a court, at the stove, behind a lens or in front of a screen.

### Board sports

- Skiing
- Wakeboarding
- Surfing

_Same rule as production: as long as it keeps moving, it stays up._

### Racket sports

- Anything played with a racket

_One request, one response. Slightly more out of breath._

### Creative

- Cooking
- Photography

_A recipe is a pipeline whose output you eat. A photo is a render you cannot re-run._

### Pop culture

- Video games
- Cinema
- And everything else

_Yes, I watch the cutscenes._

## Timeline

- 2008 — Master’s degree in communication
- 2008 — Freelance — mobile app development and DevOps
- 2016 — JobyPepper — technical co-founder, CTO
- 2023 — Ozly — technical co-founder, CTO · on hold
- 2026 — Origin AI — personal project

### Skills

- **AI & MLOps**: LLMs, RAG, Embeddings and vector search, Fine-tuning (CamemBERT), Triton Inference Server, CUDA, MCP servers, LangChain
- **Deployment**: Docker (CUDA images, GPU Compose), Kubernetes, CI/CD (CircleCI), Nginx, Sentry
- **Back end**: Node.js / Express, Python, RabbitMQ, MongoDB, Redis, WebSocket
- **Front end & mobile**: React, React Native / Expo, iOS and Android releases
- **Management**: Technical leadership (up to 10 people), GDPR, DAC7, Tax, CII

## This page is a reply.

I did not write this copy. I asked two AIs that have known me for years, then every sentence was checked against the only records that don’t lie: my private git repositories.

The original prompt (alain → claude, chatgpt):

> We’ve been talking for years. Based on our conversations and my git repos, write me the content for a landing page that shows what I can do.

- **The memory**: Years of conversations with Claude and ChatGPT: what I build, how I work, what I enjoy. Self-reported.
- **The proof**: My private git repositories: code, commits, configuration. A claim with no trace in the code does not get published.
- **The referee**: Me: the brief, the choices, the proofreading. The AIs propose, I decide.

_Yes, an AI wrote “the CTO who still commits”. It had counted the commits._

## A project, a role, a question about RabbitMQ?

The fastest way is Instagram. Based between Geneva and Lyon.

- [Instagram](https://www.instagram.com/wizopunker/)
- [LinkedIn (if you must)](https://www.linkedin.com/in/alain-lorenz-4029862b/)
- [al@inlore.nz](mailto:al@inlore.nz)
- [CV (PDF)](https://ala.inlore.nz/cv/alain-lorenz-cv-en.pdf)
- [ATS-friendly CV (PDF)](https://ala.inlore.nz/cv/alain-lorenz-cv-ats-en.pdf)

