Senior AI Engineer
Posted 4
views
Summary
Remote Senior AI Engineer who owns the AI layer of Techsa's technology and data platform: building retrieval/RAG, agent and tool-calling systems, and natural-language interfaces over enterprise data, with evaluation to measure quality. Core stack is Python and LLM application engineering, with optional vLLM serving and fine-tuning.
This is a remote position.
We are looking for a Senior AI Engineer to join our team and take ownership of key areas within our technology and data platform. Owns the AI layer: retrieval, agents, and natural language interfaces over enterprise data, with evaluation so quality is measured rather than demonstrated.
Key Responsibilities
• Own and deliver solutions within the scope of the role, from requirements and technical/design decisions through implementation and continuous improvement.
• Work closely with engineering, product, data, design, and business stakeholders to translate requirements into practical, scalable solutions.
• Apply strong engineering and/or domain expertise to build reliable, maintainable, and production-ready capabilities.
• Contribute to architecture, standards, documentation, quality, and technical decision-making appropriate to the role.
• Identify performance, scalability, data quality, usability, reliability, or operational risks and address them proactively.
• Collaborate across teams to ensure solutions integrate effectively with existing systems and platform components.
Requirements
• Experience: 5+ years of relevant professional experience.
• Strong hands-on experience with: LLM application engineering, RAG and retrieval architecture, embeddings and vector search, agent and tool calling systems, LLM evaluation, Python.
• Software or data engineering background, with at least two years building LLM based systems that reached production.
• Retrieval architecture: chunking, embeddings, vector and hybrid search, and why naive RAG fails on structured data.
• Agent or tool calling systems, including permissions, scoping, and guardrails.
• Evaluation practice: test sets, regression suites, hallucination and grounding checks.
• Comfortable working inside a data platform rather than calling a hosted API.
Preferred Qualifications
• Self hosted open weight model serving with vLLM or equivalent, inference optimisation, and GPU resource management.
• Fine tuning or adaptation of open weight models.
• Text to SQL or semantic layer work over a real data model.