Skip to content

AI Engineer at Krans

  • On-site
    • Rotterdam, Zuid-Holland, Netherlands
  • Ventures

Job description

We're building Krans, the company brain for teams that run on customer conversations. Krans connects every meeting, email, message and document a company has, and turns it into evidence: what customers actually said, which deals are really moving, and what to do next. It's live, customers have been onboarding since August, and feedback comes in directly to the founders.

We've been building fast. But we see what Krans could be, and we want to get there even faster. That's where you come in.

We're looking for the best AI engineer out there. Not a generalist who also does AI. Someone who lives in this stuff, ships fast, and wants ownership over what we build next.

How we got here

Builders is an AI venture studio in Rotterdam. We co-found AI enterprise companies with founders from day zero, which means we sit inside dozens of founder conversations, customer calls and pipelines at the same time. And we kept running into the same problem: the real picture of a company lives in meetings, inboxes and chats, and nobody has all of it. Teams run on the story they tell themselves, not on what their customers actually said.

In December 2025, Michael started building the fix himself, during his holiday. It quickly became the operating system of our own studio. In August 2026 it became a company with its own name, and we onboarded our first external customer. Krans was built inside a venture studio because nothing else worked.

Our vision is simple: every company is built on Krans. We start with company builders and portfolio holders who run multiple ventures, and we won't stop there.

What Krans actually does

  • Connects meetings, email, Slack, WhatsApp, documents and CRM into one knowledge graph per company

  • Links every claim back to the exact passage where a customer said or wrote it

  • Grades what customers say, from implied interest to real commitment, and keeps internal beliefs separate from customer statements

  • Tracks how claims change over time, so drift is visible instead of overwritten

  • Spots what's missing: unnamed budgets, unasked questions, stalled deals

  • Writes nightly company summaries you can view as of any past date, and suggests concrete next steps

  • Opens the graph to Claude, ChatGPT and Cursor through MCP: read-only, sourced and scoped by permissions

  • Gives portfolio holders one view across all their ventures

Think of it as the colleague who was in every meeting, read every email, and always knows where something was said.

What you'll work on

You'll be one of the core engineers on Krans and own large parts of the AI stack. Concretely, you'll be working on things like:

  • Extraction and synthesis pipelines that hold up across messy, multi-channel, multi-speaker, multi-language data

  • Claim extraction, grading and versioning: turning raw conversation into structured, source-linked evidence that stays correct as new information comes in

  • Agent orchestration: the agents that watch accounts, detect gaps and draft next steps, in real time

  • Multi-tenant architecture for hundreds of companies and portfolios: permissions per person and per company, strict isolation, no cross-company reads

  • The developer surface: API, MCP server and CLI, so external agents can work with Krans data

  • Integrations: new sources, more reliable sync, less noise

  • Eval infrastructure so we can actually tell when extraction and summaries get better, not just feel it

  • Latency, cost and quality trade-offs in production, not in a slide deck

You will ship to real users in week one. That's the standard.

Job requirements

Our stack (high level)

You don't need to match this exactly, the specific versions and providers will keep moving. This is the shape of what's already running for real customers.

  • Language: TypeScript end-to-end, Python where the AI tooling needs it

  • App and infra: Next.js and React on Vercel, with real-time agent orchestration and background pipelines running there

  • Data: Supabase, Postgres with pgvector, Prisma, Zod, S3

  • Models: mixed by job, currently Anthropic and Gemini, with AssemblyAI for transcription

  • Sources: Google Meet, Microsoft Teams, Zoom, Granola, Read.ai, Circleback, Slack, WhatsApp, Gmail, Outlook, Notion, Google Drive, Google Calendar, HubSpot and GitHub

  • Privacy posture: EU-hosted in Frankfurt, encrypted in transit and at rest, customer data is never used for training, permissions per person and per company. A company's context is its most valuable asset, and that shapes a lot of our architectural decisions.

Where we'd love your fingerprints: real evals around claim extraction and grading, smarter entity resolution across channels (who is who across email, Slack and meetings), the portfolio view, faster and cheaper nightly synthesis, and a developer platform other agents love to build on.

You'll probably love this if you

  • Have built and shipped real production AI systems: RAG, agents, knowledge graphs, doesn't matter, but it has to have actually run for real users

  • Have 5+ years building production platforms, ideally multi-tenant SaaS

  • Are AI-first: you use Claude Code, Cursor or similar every day and live in modern LLM and agent tooling

  • Have full-stack range, from frontend to database to agent prompts

  • Think in data models: strong relational and vector database design

  • Care about evals as much as you care about the model itself

  • Move fast, kill your darlings fast, and don't need a roadmap to know what to do tomorrow

  • Can sit with a customer, debug why a company summary got it wrong, and ship a fix the same day

Bonus points

  • You've built GTM, CRM, sales intelligence or B2B SaaS products before

  • You've built MCP servers or developer tools

  • You've worked on information extraction, entity resolution or knowledge graphs

  • You have a background in real-time transcription or speech

  • You've contributed to open source or founded a startup yourself

What we offer

  • Competitive base (€60k to €85k, depending on experience) plus equity through Builders' STAK structure. Real ownership

  • A tool budget for AI tools, IDEs and model APIs

  • Direct work with the founders, who've built companies for decades

  • A product with real customers today, no "we'll get there"

  • Weekly shipping, daily standups and Friday ship days

  • The kind of speed and access you only get this early

  • Rotterdam-based role, in our office near Central Station. We work in person, this isn't a remote position

Krans is a Builders venture.

or

Contact us

Don’t hesitate to reach out with questions, comments, or just for a quick hello.

Join our ventures

Are you looking for a great place to work, be challenged, and learn from other bright minds?