Founding AI Engineer @ Embedded AI Efficiency Layer, Berlin

atlantic.vc; Berlin; Vor Ort; EUR 90.000 - 130.000; Vollzeit

Zusammenfassung

Atlantic is seeking an AI engineer to join early and help build the core product from the ground up. You will define architecture, own major technical directions, and turn research ideas into reliable production systems across AI infrastructure and enterprise deployments.

You will work with founders on token reduction, model routing, and secure enterprise infrastructure, with strong ownership of design choices and deployment for high-scale production.

Qualifikationen

• Candidates should have strong software engineering fundamentals and production AI experience.
• Experience with backend systems, API development, and cloud infrastructure.
• Fluency in English; German is a plus.

Aufgaben

• Research and productionise AI cost-efficiency systems for context optimisation.
• Develop context selection, compression, caching, model routing and output budgets.
• Build APIs, an OpenAI-compatible gateway and provider integrations.
• Create reproducible cost-quality evaluations, regression tests and quality gates.
• Own production reliability across monitoring, deployments, incidents, capacity and latency.

Kenntnisse

Python

TypeScript

API development

Distributed systems

Docker

Kubernetes

PostgreSQL

Cloud

English

German

Ausbildung

Bachelor's degree in CS or related

Tools

Docker

Kubernetes

PostgreSQL

Jobbeschreibung

This is an Atlantic portfolio venture. About the Venture AI is being integrated into more products every month, and every interaction consumes tokens. As usage scales, costs add up quickly. The first race was adoption. The next is efficiency: inference costs increasingly determine what companies can afford to build and ship. We're building the efficiency layer for production AI. Our systems make AI dramatically leaner, reducing token usage and lowering the cost of every interaction without compromising the quality standards that matter. Over time, we aim to improve both cost and quality, so companies can do more with AI rather than choose between better performance and a healthier bottom line. We closed our pre-seed in just three weeks and are now preparing our first customer pilots. The pace is high and the timelines are ambitious. We are aiming for something big, and that requires a lot of input from every member of our early team.

About the Role

You'll join as our first AI engineer and work directly with the founders to build the core product from the ground up. You'll own major areas of product development, help define our architecture and technical direction, and turn complex research and engineering problems into reliable production systems. This role requires someone who can think deeply about AI systems and write the code to build them. You'll work across AI infrastructure, backend systems, evaluation and enterprise deployments, with substantial ownership over the technical decisions behind the product.

What You'll Do

• Research, invent and productionise systems for AI cost efficiency, context optimisation and token reduction.
• Build across context selection, compression, caching, model routing, tool-call reduction, retry control and output budgets.
• Develop APIs, an OpenAI-compatible gateway and integrations with model providers and enterprise AI applications.
• Create reproducible cost-quality evaluations, regression tests and quality gates.
• Own production reliability across monitoring, deployments, incidents, capacity, availability and latency.
• Build secure enterprise infrastructure with tenant isolation, access controls, secrets management and protected customer data.
• Optimise distributed-system performance across throughput, caching, rate limits, resource usage and failure handling.
• Translate early customer requirements into reliable product capabilities.
• Take on the broad day-to-day engineering work required to move an early product forward.

About You

• You're an experienced AI engineer who combines strong software engineering with original thinking about how production AI systems can become more efficient.
• You can reason from first principles about where cost and quality are lost, test new approaches rigorously and turn the strongest ideas into production-grade code.
• You have substantial experience in software engineering, backend systems and API development.
• You are highly proficient in python or TypeScript or at least some coding languages and comfortable working across them.
• You understand the foundations of modern LLM systems, including tokenisation, context management, inference behaviour, model evaluation and the trade-offs between cost, latency and quality.
• You have hands-on experience building with LLM APIs and understand the behaviour, cost and reliability challenges of production AI systems.
• You have practical experience designing experiments and evaluations that measure both model quality and system performance.
• You have worked with cloud infrastructure, PostgreSQL and Docker.
• Kubernetes experience is a strong advantage.
• Experience with vLLM, SGLang, LiteLLM or similar AI infrastructure is highly relevant.
• You understand distributed-system performance, including latency, throughput, caching, rate limits and failure handling.
• You can independently research difficult problems, test hypotheses and translate technical ideas into working systems.
• You write clear, reliable code and are comfortable owning systems through deployment and production.
• Fluent English is required.
• German is an advantage.

We do not expect you to have worked on every system or technology listed above. We do expect deep experience in some of these areas, strong software engineering fundamentals and the ability to develop genuine technical depth in the rest.

Why Join

You'll join at the earliest stage and help shape the product, architecture and company strategy. You'll invent core technology directly with the founders and receive meaningful equity in what you help build. The pace is demanding and the technical bar is high, but the scope and opportunity are significant. This role suits an AI engineer who wants substantial ownership and the chance to work on difficult, commercially important infrastructure problems.

This is a full-time role ba