CTO Relief

TL;DR

Title Co-Founder & CTO

Location Paris, 75003 — in person

Equity Late founder structure — to negotiate with Mickaël

Cash Competitive post-raise salary, to discuss

Funding €2M in the bank. 12–15 months runway. Next raise in preparation.

Start ASAP

🌍 What Relief does

Relief is building the reference quoting platform for manufacturing running custom, project-based work.

These companies quote with Excel or intuition, with no memory of past jobs and no AI assistance. Each quote takes hours. Experienced estimators are scarce. Business is lost because response time is too slow.

Relief fixes that. Our AI assistant reads technical plans, structures a full quote, and dramatically reduces quoting time — enabling clients to quote 3–5x faster, answer more RFQs with the same team, and capture business they previously had to decline.

We start with metal sheets and tubes workers (tôlerie-chaudronnerie) because it is the most acute expression of this problem — high complexity, scarce expertise, and strong urgency to quote faster. Crack it here, and we have the proof to scale across multiple manufacturing segments in Europe.

The deeper strategic bet

ERPs own production. CRMs own contacts. But everything before production starts — the RFQ, the cost assumptions, the estimator's tribal knowledge — is never structured, never stored, never reused. It evaporates.

Relief captures this layer for the first time. Quoting is the entry point because it's where the data originates. Over time, Relief becomes the system of record for commercial intelligence that neither ERP nor CRM will ever own — sitting between the two, owning what neither does.

🎬 Where we are — and why we're here

Relief raised €2.6M, signed paying clients across multiple manufacturing segments, and built a first version of the product. Then we hit a wall.

We made wrong architectural bets: putting AI where determinism was needed, building agentic chains where a single model call would have done the job. We accumulated organizational debt in the tech team. We tried to cover too many segments at once.

So we made hard decisions:

• Let go of 50% of the team

• Terminated most client contracts

• Narrowed the entire company to one segment: tôlerie-chaudronnerie, for the next 6 months
  
  We now have a crystal-clear hypothesis: if we can make metal sheet workers' quoting dramatically faster and more accurate, we have an unfair advantage to scale Relief across geographies (Europe and the US) and adjacent segements. The next version of the product ships in 2 months. First clients are co-building it with us.
  This is not a "fire and forget" pivot. It's a deliberate, focused rebuild — with enough cash to execute and a clear path to Series A if we validate the hypothesis by September.

🧭 What you'll do

You take full technical ownership of Relief alongside Mickaël.

You audit what was built, decide what to keep and what to rebuild, and set a new technical direction — one grounded in the actual problem, not the previous team's assumptions. You make architectural calls that compound: right now that means a system that reads technical plans reliably, structures quotes accurately, and integrates with the messy reality of SME data environments.

You bring AI into the product the right way — not as a marketing layer, but as the core of the value engine. You know when to use a single model call versus a pipeline, when determinism beats probabilistic generation, and how to evaluate model output in production with real users. Relief's AI needs to be reliable, configurable, and measurable. You make that happen.

You lead the team. You raise the bar. You build a technical organization that ships at a predictable pace without depending on you for every decision.

You're Mickaël's technical sparring partner — on product decisions, architectural tradeoffs, client conversations, and investor meetings. When we raise Series A, you're in the room, you own the technical narrative, and you defend every architectural choice with facts.

You run client calls. You understand that in a B2B industrial context, the best technical decisions come from sitting in the room with estimators who actually use the product. You don't hide behind Slack.

🧱 The tech team today

3 FTE:

• 1 AI Engineer

• 2 Lead Software Developers (10 years XP each)

No CTO. No architectural debt meetings. No politics. Just execution waiting for direction.

📊 What success looks like

30 days

Full technical audit. You know what to fix, in what order, and why. At least one thing shipped.

90 days

The team ships reliably. AI architecture is defined, documented, and in execution. You've been in at least 3 client conversations.

6 months

Relief is technically defensible to Series A investors. The system is measurable, stable, and showing clear improvement curves. Mickaël has a real technical co-founder.

🧠 Who you are

You've been a technical co-founder in an early-stage context — real equity, irreversible architectural decisions under pressure, full ownership of what shipped and what didn't. You know what it costs to get the architecture wrong at 10 people.

You've transformed complex physical data into reliable, structured systems — in production.

Not research. Not demos. Production, with real users, heterogeneous inputs, and consequences when the output is wrong — document understanding, computer vision on technical files, or structured extraction from unstructured sources.

You know where to use AI and where not to.

You have a hard-earned point of view on the boundary between probabilistic and deterministic systems. You've seen what happens when you put a language model where a rule engine belongs — and you've fixed it.

You think in domain models, not just data models.

You've structured complex domain knowledge into something a machine can reason about durably. You know the difference between an ontology that holds for 12 months and one that breaks at the first edge case.

You build engineering organizations that don't depend on you for execution. You set standards, grow the people who are already there, and trust the team to ship.

You're direct and structured. You turn ambiguity into a plan, decisions into documented outcomes, and hard conversations happen early — with your co-founder, with clients, with investors.

Experience required:

• 3+ years as Co-founder CTO in an early-stage context — real equity, real ownership

• Proven experience deploying AI on complex, heterogeneous physical or document data — in production

• B2B SaaS with real data integration constraints

Strong plus: computer vision on technical documents, industrial data environments, ERP-adjacent systems, or domain modeling in manufacturing or engineering contexts.

🚫 This role is probably not for you if…

• You want a stable scope

• You think of AI as a feature layer

• You avoid difficult conversations

• You need a large team to be effective

• You're looking for financial security without the asymmetry of equity

• You don't want to talk to clients

💡 Why join Relief

• Real traction, clear focus. We're not searching for a market. We've found it, proved initial demand, and are now executing with precision.

• Technical leadership from day one. No politics, no legacy CTO above you. Full ownership, full responsibility.

• A founder who can sell. Mickaël has closed clients, investors, and key partnerships. The commercial half is working.

• A problem worth solving. Manufacturing companies in Europe & US are 10 years behind on tooling. The gap is enormous.

• Meaningful equity. Late founder structure, negotiated with Mickaël. Real upside if we win.

External validation. Relief was selected in the Notion Capital Cloud Challengers 2026 — top 100 European B2B software companies to watch — and named amongthe top 20 most promising French startups by 20VC. Two independent signals that the thesis is real.

💼 The package

Equity

Meaningful late-founder structure — negotiated with Mickaël

Salary

Competitive, post-raise, based on profile

Office

Paris 75003 — in person expected

Ecosystem

OSS Ventures portfolio network, shared technical resources, product and revenue team