AI Engineer

tl;dr

• After 14 years in construction tech, Lea & Jonas are building the intelligence layer for the construction industry

• The largest construction enterprises in Europe with a combined project volume of over €30bn annually rely on Kuro every day to automate workflows and make faster, better decisions

• Construction has the most complex, unstructured data you can throw at an LLM — PDFs, drawings, 3D models, images, calculations. At Kuro you can tackle the hardest applied AI problems everyday.

Lea and Jonas scaled two of Europe's leading construction tech companies
- Capmo and alcemy - and saw firsthand the challenges holding the industry back. Now they're tackling the root cause: the 50% of construction work still lost to manual data handling.

Kuro is the the intelligent project data system for construction. We're reimagining how construction professionals get work done and make decisions by leveraging AI to augment back-office workflows and make core data accessible.

After one year on the market leading construction enterprises with a combined construction volume >30 billion annually use Kuro everyday to automate workflows and make faster, better decisions.

We are 14 people based in Berlin. You will join a group of ex-technical founders and builders that are ambitious and kind.

Learn more:
Join Kuro
- Berlin

As AI Engineer
, you'll work directly with Kuro’s CPO Jonas on building and scaling high-impact AI automation use cases for some of the most complex construction projects in the industry. This is a highly hands‑on and high‑ownership role at the intersection of AI, product, and customer impact, where no two projects look the same. Each customer project consists of hundreds to thousands of files — including PDFs, technical drawings, scans, and other unstructured data sources — and your role will be to prototype, optimize, and product ionize AI‑driven workflows and automation use cases that save customers significant amounts of manual work and unlock measurable operational impact.

This is the perfect role for an experienced engineer who wants to join an early stage company that builds a high impact applied AI product - for someone who enjoys solving real-world problems with AI, thrives in fast-moving environments, and wants to work closely with both product and customers in an early-stage company. You’ll take ownership from day one, experiment rapidly, and help turn ambitious ideas into scalable product capabilities.

Optimize and Productionize Solutions

• Manage experimentation pipelines and perform error analysis.

• Oversee evaluations, scoring, and golden data set creation.

Enhance ML and Data Pipelines

• Continuously improve pipelines to support new and existing use cases

Refine Evaluation Processes

• Optimize evaluation and scoring systems for robust performance monitoring.

• Boosted accuracy by 9% through prompt engineering with customer feedback.

• Achieved an 11% accuracy improvement using LLM for prompt rewriting (LLM-as‑a‑audience).

• Reduced hallucinations by implementing an LLM-as-judge approach.

• Enhanced search relevance through BM25/Hybrid Search implementation.

• Developed use cases like mapping 3D-model metadata using LLMs.

It is still day 0 for Kuro. Therefore we are highly pragmatic with our tool choices. We believe in keeping it simple and moving quickly. You significantly help shape our tech stack decisions.

• Web-App:
  Typescript / Vite

• Cloud Functions:
  Supabase

• Database / Vector DB and File Storage:
  Supabase (postgress) / Weaviate / AWS S3

• Backend (Document Pipelines & applied AI):
  Fastify API / TS + Python for ML Services

• Hosting:
  Railway (Moving to AWS)

• AI:
  Azure / OpenAI, Gemini, Cohere

• Obsv., Eval. & Tracking:
  Langfuse, Datadog

• You thrive working in a small and accountable team with a lot of ownership

• You love 0 to 1 phase

• You want to take full ownership in solving customer problems by designing and implementing features

• You are passionate about the potential of GenAI, a strong believer in its transformative power, and eager to explore how it can be applied to solve real-world challenges in the…