Senior AI Engineer

Ersilia; Hofheim am Taunus; Remote; EUR 90,000 - 130,000; Full time

Job summary

Ersilia is seeking a hands-on Senior AI Engineer with GenAI experience to move fast, test aggressively and scale what works. You will own end-to-end solutions, work with leadership, and help drive our next growth phase.

You will implement agent features, integrate LLMs, and build reliable autonomous agents with observability and governance in enterprise environments.

Qualifications

• 6+ years in software engineering with 3+ years in GenAI for production systems.
• Proficient in Python with production deployment experience.
• Strong AWS knowledge (CDK, Serverless, Bedrock) with production deployment experience.
• Hands-on with LLM providers and frameworks: Azure OpenAI, AWS Bedrock, Hugging Face, LangChain/LangGraph.
• Deep expertise in agent frameworks and the ability to build custom orchestration.
• University degree or equivalent experience in Computer Science, Engineering, or a related field.

Responsibilities

• Implement agent features and integrations in production codebases – APIs, websockets, tool-calling workflows – following established architecture patterns.
• Adapt and integrate pre-trained LLMs and embeddings into applications (prompting, fine-tuning, RAG pipelines) and keep them maintainable.
• Build self-verifying agent workflows with chain verifications, evaluator models and human looped escalation paths that make agents trustworthy in enterprise environments.
• Create comprehensive agent observability, full audit trails, structured logging, performance metrics – so the team always has clear agent visibility.
• Ship autonomous agents making real business decisions with high reliability, optimising the balance between reasoning depth, latency, and cost.
• Embed with product teams to evolve agent capabilities based on real usage data, closing the loop between what’s built and what actually works.

Skills

GenAI experience

Software engineering

Agent frameworks

Education

University degree in Computer Science or related field

Tools

AWS CDK

Serverless

Bedrock

Azure OpenAI

Hugging Face

LangChain

LangGraph

Job description

We are looking for a hands-onSenior AI Engineer with GenAI experiencewho can move fast, test aggressively and scale what works. You will own the end-to-end solution, work closely with leadership and play a key role in our next phase of growth.

Your responsibilities:

• Implement agent features and integrations in production codebases – APIs, websockets, tool-calling workflows – following established architecture patterns.

• Adapt and integrate pre-trained LLMs and embeddings into applications (prompting, fine-tuning, RAG pipelines) and keep them maintainable.

• Build self-verifying agent workflows with chain verifications, evaluator models and human looped escalation paths that make agents trustworthy in enterprise environments.

• Create comprehensive agent observability, full audit trails, structured logging, performance metrics – so the team always has clear agent visibility.

• Ship autonomous agents making real business decisions with high reliability, optimising the balance between reasoning depth, latency, and cost.

• Embed with product teams to evolve agent capabilities based on real usage data, closing the loop between what’s built and what actually works.

What we are looking for:

• 6 years+ in software engineering with 3+ years in GenAI, building and shipping production systems in enterprise settings.

• Strong proficiency in Python with production deployment experience – you write code that other engineers can maintain, test, and extend.

• Solid AWS knowledge (CDK, Serverless, Bedrock) with production deployment experience.

• Hands-on experience with LLM providers and frameworks: Azure OpenAI, AWS Bedrock, Hugging Face, LangChain/LangGraph.

• Deep expertise in agent frameworks combined with the ability to build custom orchestration.

• University degree or equivalent experience in Computer Science, Engineering, or a related field.

This is a fully remote position which can be based globally given a +-4 hour time difference within the CET region.