Software Engineer - Artificial Intelligence/LLM

Description :

- Design, develop, and deploy LLM-based applications for clinical and operational workflows (summarization, chat, reasoning, data extraction).

- Implement and optimize RAG (Retrieval-Augmented Generation) pipelines with vector databases (FAISS, Pinecone, or Vertex AI Matching Engine).

- Work hands-on with LLM APIs (OpenAI, Claude, Gemini, Mistral, Llama) for fine-tuning, evaluation, and productionization.

- Build and maintain AI microservices / APIs using FastAPI or Flask, integrated with Folds Unity platform.

- Implement Agentic AI workflows (multi-step reasoning, tool use, and task orchestration) using frameworks such as LangGraph, CrewAI, or Semantic Kernel.

- Collaborate with the Data Engineering and Product teams to ensure seamless integration of models with curated EHR and FHIR data.

- Conduct model evaluation and benchmarking (factual accuracy, hallucination control, latency).

- Follow HIPAA-compliant development practices and ensure PHI safety and responsible AI usage.

Must have Skills :

Python :

Experience in Genarative AI :

Experience in FHIR, HL7 :

- Strong Python skills with experience in LLM frameworks (LangChain, LlamaIndex, Hugging Face Transformers) preferably from Healthcare domain.

- Proven experience building production-grade LLM or Agentic AI applications (e.g., chat assistants, document agents, summarizers, copilots).

- Proficiency in RAG architectures, vector search, and embeddings.

- Experience working with cloud AI platforms (Vertex AI, Azure OpenAI, AWS, Bedrock, or Anthropic Console).

- Knowledge of Agentic AI use cases autonomous task handling, reasoning chains, planning-execution loops.

- Understanding of healthcare data formats (FHIR, HL7, clinical notes) is highly preferred.

Good to have Skills :

- Knowledge of MLOps, continuous evaluation pipelines, or model observability.

- Experience integrating AI features into SaaS or API-first products.

- Prior exposure to healthcare AI, clinical NLP, or population health analytics.