Senior AI Engineer

What you'll work on

You will design and build AI systems for patient identification, clinical trial feasibility, recruitment automation, and agentic workflows.

This includes:

• Building RAG and agentic AI systems over structured and unstructured clinical data.
• Designing workflows for eligibility reasoning, patient matching, recruitment automation, and recruiter assistance.
• Developing large-scale data ingestion and processing pipelines across EHR data, FHIR resources, PDFs, clinical notes, and trial criteria.
• Building embedding, indexing, retrieval, reranking, and evaluation pipelines using tools such as Milvus/Zilliz, Elasticsearch, and domain-specific embedding models.
• Improving the quality, reliability, latency, and cost-efficiency of AI workflows in production.
• Working with open-source and hosted models for extraction, classification, reasoning, summarisation, and patient-trial matching.
• Designing evaluation frameworks to measure accuracy, retrieval quality, hallucination rates, and overall system performance.
• Helping build infrastructure capable of processing millions of patient records and large volumes of healthcare data.
• Improving prompt design, tool usage, orchestration logic, and system robustness.

What we're looking for

We are looking for someone who has worked in a startup or high-ownership environment before.

You should have strong engineering fundamentals and a deep understanding of how modern AI systems behave in production including where they fail.

You should be someone who can think beyond “calling an LLM API” and understand the broader challenges of building reliable AI systems: retrieval quality, context construction, evaluation, latency, model selection, data quality, orchestration, cost, and observability.

Must Have

• Strong hands-on experience building production AI applications using LLMs, RAG, agents, or related architectures.
• Experience shipping and maintaining production-grade AI systems, not just prototypes.
• Strong software engineering skills in Python, Node.js, or similar languages.
• Experience with large-scale data processing, ingestion pipelines, ETL/ELT, or workflow orchestration.
• Experience with vector databases, embeddings, search, indexing, retrieval, and ranking systems.
• Understanding of prompt design, structured outputs, tool calling, and multi-step AI workflows.
• Ability to work with messy real-world data and design robust, scalable pipelines.
• Experience working in a startup, early-stage company, or similarly fast-moving environment.
• Ability to take ownership of problems end-to-end, from architecture and implementation through deployment, monitoring, and iteration.

How We Work

We are a small, high-ownership team building in a complex and meaningful space.

You will be expected to move fast, but with discipline. Clinical research is not a space where flashy demos are enough. Accuracy, reliability, auditability, privacy, and operational usefulness matter.

We value people who:

• Think from first principles.
• Care deeply about product and customer impact.
• Can operate with ambiguity.
• Are comfortable taking ownership without waiting for detailed instructions.
• Challenge ideas respectfully.
• Ship fast, learn quickly, and improve continuously.
• Understand that building real AI products means dealing with messy data, edge cases, latency, costs, and failure modes.

This is a role for someone who wants to build at the frontier of applied AI in healthcare.

Compensation

We offer a competitive salary package with meaningful equity options.

For the right person, this role is designed to be a long-term, high-impact position within the company.

You will be joining at an important stage of growth, with the opportunity to shape the AI architecture, technical direction, and product intelligence layer of Srotas Health.

NOTE:

Please share your CV, GitHub or portfolio if available, and a short note on one production AI system or data pipeline you have built that you are proud of.