Senior AI Engineer

An AI consulting group that builds serious production-grade systems for companies ready to operationalize artificial intelligence across their business. We do not build gimmicks. The main focus is on custom AI solutions that integrate directly into operations and drive measurable results.

Were looking for a Senior AI Engineer for one of our clients an AI consulting group that builds production-grade systems for companies ready to operationalize artificial intelligence across their business. This is a hands-on role for someone who doesnt just experiment with AI but has built deployed and optimized real AI systems in complex production environments.

This role is for someone who takes full ownership of AI systems from architecture to performance. Youll work at the intersection of LLMs distributed data processing and backend systems building scalable AI pipelines and knowledge systems in a high-impact environment where performance reliability and flexibility (including airgapped deployments) matter.

Your responsibilities will include:

• Design and build distributed data pipelines (Apache Beam Python) for large-scale document and model processing;
• Implement horizontally scalable LLM-powered extraction workflows across thousands of entities;
• Build runner-agnostic pipelines that run across multiple environments (local cloud on-prem);
• Design and implement entity resolution and cross-referencing systems with confidence scoring;
  Architect and build graph-based Knowledge Bases (Dgraph or similar) including schema design and query optimization;
• Develop pluggable LLM extractor frameworks (classification relationships completeness document linking);
• Build abstraction layers for LLM providers (Anthropic Claude Google Gemini local models via Ollama/vLLM);
• Optimize LLM usage for cost latency and reliability in production environments;
• Design and implement backend service layers (Go-based APIs) connecting AI pipelines with agent systems;
• Ensure observability debugging and performance monitoring across distributed AI systems;
• Package and deploy systems using Docker (docker-compose containerized environments);
• Collaborate directly with client teams on integrations validation and production rollout.
  
  

What we expect from you:

• 58 years of experience in AI Engineering Data Engineering or Applied ML Infrastructure;
• Expert-level Python;
• Strong experience with distributed processing frameworks (Apache Beam preferred; Spark/Flink acceptable);
• Hands-on experience integrating LLMs in production (prompting orchestration provider abstraction);
• Experience with graph databases (Dgraph Neo4j JanusGraph or similar);
• Strong understanding of distributed systems parallel processing and large-scale data pipelines;
• Solid SQL and data modeling fundamentals;
• Experience with Docker and containerized deployments;
• Experience with cloud platforms (GCP preferred AWS/Azure acceptable);
• Strong backend/system design mindset (performance scalability reliability);
• Ability to work independently and take ownership without micromanagement;
• Strong problem-solving skills in complex ambiguous environments;
• English level: B2 (C1 preferred);
• Availability to overlap with US business hours.
  
  

Nice to have:

• Working experience with Go (for API/service layer contributions);
• Experience with local LLM inference (Ollama vLLM TensorRT-LLM);
• Experience deploying systems in airgapped or restricted environments;
• Background in defense aerospace or complex enterprise systems;
• Familiarity with MBSE / SysML tooling and model-based data (XMI Cameo etc.);
• Experience with document parsing pipelines (Docling Unstructured etc.);
• Experience with orchestration tools (Airflow Dagster Prefect).
  
  

We offer:

• Four-month full-time contract engagement;
• Remote work with overlap in US business hours;
• High-impact project with direct influence on production AI systems;
• Close collaboration with clients product and engineering teams;
• Opportunity to build core infrastructure for real-world AI applications not prototypes.