Founder-led · Hannover · Remote-first
Forge Labs UG builds full-stack apps, MVPs, AWS infrastructure, workflow automation, and AI features for small teams in Germany. Validated, documented, and built so your team can own it after handover.
What Forge Labs does
Not another chatbot on top of a model. Forge Labs builds structured, auditable systems and the engineering practices that keep them reliable.
We combine full-stack product engineering, AWS architecture, and applied GenAI into software that can actually be used.
Who it's for
Turn an idea, roadmap pressure, or feature gap into a usable MVP, SaaS feature, or clear technical plan.
Turn manual, error-prone work into validated workflows with logs, review points, and clear failure paths.
Add senior GenAI, RAG, AWS, or full-stack capacity for scoped delivery without building a new internal team.
Services
React/Next.js, TypeScript, Java/Kotlin, Spring Boot, APIs, tests, and CI/CD.
You get a usable app, MVP, or feature slice your team can extend.
Multi-step LLM workflows with structured outputs, validation, retries, and audit trails.
You get automation for work that is slow, manual, or error-prone today.
Grounding, vector search, citations, evaluation, and safe UX for internal or user-facing assistants.
You get AI answers that are traceable, testable, and reviewable.
CDK, DynamoDB, Lambda, ECS, SQS/SNS, monitoring, right-sizing, and cost review.
You get less cloud waste, lower risk, and a prioritized implementation plan.
Cursor, Claude, Copilot, custom agent workflows, team rules, review processes, and handover docs.
You get AI as a real engineering lever, not prompt chaos.
Engagements
No endless consulting decks. Pick the format that fits — pricing is scoped after a short discovery call.
A defined deliverable — an MVP, a GenAI pipeline, a RAG system, or an AWS cost audit. Scoped, built, tested, and handed over so your team owns it.
Ongoing senior architecture, implementation, and AI enablement — the senior capacity of a hire, without the full-time commitment.
Embedded with your team for a defined period to ship one specific, well-scoped outcome end to end.
Crew Orbit & engineering
Forge Labs is not an anonymous agency. The work comes from Kevin Voß's experience as a full-stack, AWS, and GenAI software engineer and from building Crew Orbit.
The AI engineering team-in-a-box. Stop prompting. Start shipping. Crew Orbit replaces prompt chaos with structured workflows for planning, implementation, validation, and review.
Senior full-stack and GenAI software engineer from Hannover focused on React/TypeScript, Java/Spring Boot, AWS, RAG, LLM workflows, and AI-assisted engineering processes.
Skills & stack
One engineer who takes a product from architecture to AI to cloud — no handoffs, no translation loss between separate teams.
Proof from production
Founder experience from SaaS, cloud, and GenAI delivery. Numbers are context-dependent, but they show what the work optimizes for: time, cost, reliability, and ownership.
AWS cost reduction through indexing, right-sizing, monitoring, and auto-scaling in a SaaS platform.
Faster delivery through architecture, reviews, testing, CI/CD, and structured engineering practices.
Manual-time reduction in a GenAI pipeline: 30-60 minutes to under 1 minute.
RAG validation, citations, and file-structure optimization for enterprise knowledge workflows.
Selected project experience includes GenAI pipelines, RAG validation, real-time infrastructure, and SaaS platforms for companies such as GS1, Siemens, Mercedes-Benz, and Micromerce.
How we work
Forge Labs is not a large dev shop with a bench. It is founder-led delivery for a small number of well-scoped engagements.
Clarify the goal, users, constraints, current stack, and first useful outcome.
Define architecture, risks, data flow, scope, and realistic calendar timing.
Ship increments, test, review, and document decisions as the work moves.
Deliver handover docs, runbooks, and next steps so your team owns the result.
FAQ
Let's build
Send a short note with what you want to build or improve: goal, current stack, first useful outcome, and timing. You will get a clear recommendation on what should happen first.