Artificial Intelligence Engineer

Seamflow; Fatih; On-site; TRY 7,843,000 - 11,765,000; Full time

Benefits offered by this job

Relocation support

Visa sponsorship

Lunch & dinner covered

Gym access

Tooling and resources

Job summary

Seamflow seeks an Applied AI Engineer to design end-to-end AI systems for the TIC industry. You will own the AI surface area, including prompts, agents, and tool orchestration, and you will engage with customers to ensure reliability and fast delivery.

You will read real documents, ship features to production in week one, and focus on robust evals to prevent regressions in regulated domains.

Qualifications

• Experience shipping AI systems to production with real users.
• Ability to diagnose model failures and explain fixes.
• Develop evals that catch regressions, not vanity metrics.
• Fast, opinionated shipping while avoiding unsafe AI behavior.
• Read customer documents and join calls to understand real outputs.

Responsibilities

• Own AI surface area end to end: prompts, agent design, tool usage, and evaluation.
• Ship to production in the first week and monitor performance.
• Read hundreds of real outputs and debug edge cases with customers.
• Collaborate with customers to understand workflows and enforce reliability.

Skills

Applied AI

Model behaviour

Evals & iteration

Agents & tooling

Document intelligence

Product judgement

Tools

Python

TypeScript

React

OpenAI SDKs

Anthropic SDK

Azure OpenAI

Job description

Sponsorship is available; relocation, visas, meals and tools are covered.

Seamflow builds AI-native tools for the TIC industry, turning complex regulatory workflows into scalable operations. Scheduling and resource planning, technical review, conformity assessment, field inspection, and document intelligence over messy real-world files.

Software moves fast, but industries like aerospace, healthcare and energy are still slowed by outdated, manual systems. We are building a system that lets complex, real-world work move at software speed.

We're a venture backed stage start up that builds AI native tools for the Testing, Inspection, and

TIC is a $300bn industry that verifies whether products, processes and systems meet specific safety, quality and regulatory standards. Currently it is a very slow and manual industry, where the opportunity for AI is huge.

We are already working with the major TIC players, experiencing rapid revenue growth and are on route to a Series A round.

The team includes former unicorn founding engineers such as Fuse Energy and people from X,

Google, Amazon and Yandex.

Commercially the team includes one of the youngest Associate

Partners at McKinsey, managers from Bain, and individuals on Forbes 30u30.

High standards, high trust, no politics. A team that wants to win and thrives on ownership.

The role

An Applied AI Engineer here thinks in workflows and model behaviour, not benchmarks or isolated prompts. The inputs are genuinely hard: 400-page technical files, scanned forms, tables that were never meant to be parsed, documents where a single misread classification has consequences. The bar is not a good demo. The bar is a system a reviewer trusts on a Tuesday afternoon with a deadline.

You own AI surface area end to end: what the system should do, the prompts, the agent design and tool use, the evals that catch regressions before customers do, the monitoring, and the iteration after it's live. You'll sit in customer calls, read the real documents, and look at hundreds of real outputs, because that's where the failure modes actually are. You'll ship to production in your first week.

What we're actually screening for:

• You've run AI systems in production, not demoed them. Something real users depended on, that broke in ways a benchmark never predicted, and that you fixed.
• You can say why a model failed. Not that it hallucinated. Why, and what you changed.
• You build evals that catch regressions, rather than metrics that flatter the system.
• Speed with taste. You ship fast, and you're opinionated about what AI should not do and when good enough ships.
• You go to the source. You'll read the customer's actual documents, join the actual call, and look at hundreds of real outputs.

You will fit if you have experience with

• Applied AI: LLM-powered features that solve real workflows, from prompt to production.
• Model behaviour: understanding why models fail, which model to use, and how to make them reliable.
• Evals and iteration: evals that catch real regressions rather than vanity metrics.
• Agents and tooling: multi-step agents, tool use and orchestration for complex domains.
• Document intelligence: structured data out of messy PDFs, scans, forms and tables.
• Product judgement: a clear view on what AI should and should not do.

Technical must-haves

• Strong Python and modern LLM tooling (OpenAI, Anthropic and Azure OpenAI SDKs, structured outputs, function calling).
• Production AI systems that go beyond a wrapper around a chat API.
• Comfort with evals, prompt engineering and model failure modes.
• Ability to work across the stack: TypeScript and React on the frontend, Python on the backend.
• Reasoning about complex domain workflows, not just prompt and response.

First month

• Ship an AI feature to production in the first week.
• Own problems end to end, from prompt to eval to deployment.
• Read hundreds of real outputs, debug edge cases and talk to customers.

Why candidates say yes

• They ship to production from day one and own real, user-facing surface area.
• The role is in person in London, with relocation, visas, meals and tools fully covered.
• The problems are genuinely hard and grounded in regulated industries, not another CRUD app.

What you get

• Real equity, and compensation that reflects a high bar
• Genuine ownership of product surface area from day one, with the founders in the room
• Relocation support and visa sponsorship, handled properly
• Lunch and dinner covered, 24/7 gym access, and any tool that makes you faster
• A small team with high standards, high trust, and no politics

Your point of contact

Ahmet Utku Yavuz, Talent Partner, Zen Talent.