Forward Deployed Construction Engineer

_Ingénieur Construction déployé chez le client (H/F)_

_Part-time to start · Remote, France (onboarding in Stockholm) · Construction × AI_

Fluent English is required for this role: it’s our working language across the company. Version française en fin de document.

About Brickanta

Brickanta is the AI-native agentic platform for construction. Hundreds of AI agents take on the administrative heavy lifting across bidding, procurement, compliance and quality assurance, so builders finish projects on time and on budget.

We’re backed by Y Combinator, have raised over $8M from Northzone, Anton Osika and Mario Götze, and already work with some of the largest contractors and developers in the industry. We’re an early team and growing fast, operating like an elite sports team: maximum freedom paired with maximum responsibility. You own your work, you document what matters, and you win and lose as a team.

Having built the platform with builders in the Nordics, we’re now taking it across Europe. France is one of our first markets, and this is one of our first roles on the ground.

The role

Most enterprise AI dies in the proof of concept. Forward deployed engineering is how we avoid that: you sit inside the customer’s workflow — on site, in the bid room, in the procurement process — and build until the work actually gets done differently.

In a new market, that also means teaching our agents how construction works here. An answer can look plausible and still be wrong in the way that matters on a real project: a CCAP clause read the wrong way, a quantity that doesn’t match the DPGF, a DTU requirement that gets missed. Your job is to spot that and turn the judgment into evals our systems can measure.

This is not consulting and it is not a demo role. You ship product: take a real bottleneck at a real contractor, build the agent workflow that removes it, prove the outcome in numbers, and bring what you learned back into the core platform.

What you’ll own

_Deployments_

Own named accounts in France end to end: understand the workflow, configure and extend our agents against their real documents and data, and get them into daily use.

_Evaluation and ground truth_

Run our agents against real French workflows and judge the outputs on correctness, completeness and practical usefulness, not on whether they sound convincing. Turn edge cases and production failures into repeatable evals: reference outputs, scoring rubrics, failure taxonomies and expert-reviewed datasets.

_Building in the field_

Write code. Integrations, data pipelines, prompts, eval harnesses, automated graders and occasionally a whole new feature: whatever it takes to make the workflow work at that customer. Work out what needs human judgment, then engineer the rest away.

_Measurable outcomes_

Define what success means before you start, then prove it: hours saved, bids won, risks caught. Measurable results rather than technical milestones.

_Feeding the product_

Capture what only shows up in production here, from DCEs and DPGFs to how tendering and procurement really run, and turn it into datasets, reusable building blocks and roadmap. Each deployment in France should make the next one faster.

Who we’re looking for

Some combination of construction depth, technical depth and a liking for customers. The unusual profile matters more than the perfect CV.

_Background_

Real exposure to construction in France: site work, works management (conduite de travaux), cost estimating (études de prix), quantity surveying (métrés), procurement or quality. Experience building and shipping software (Python and/or TypeScript, APIs, data pipelines, SQL), or the raw ingredients and the drive to get there quickly. Strong analytical ability and a serious interest in modern AI systems. We operate as an AI-first company. Experience working directly with customers, or a clear appetite for it. Fluent English, since it’s our working language, and native or near-native French. N