Artificial Intelligence Engineer

Role & responsibilities

• Build, test and deploy Python-based SDKs, APIs and microservices for generative/agentic workflows: retrieval-augmented-generation (RAG), multimodal text/chart/table ingestion, conversational human-assistant dialogues.
  
• Implement, fine-tune and optimise machine-learning / deep-learning models: supervised fine-tuning of LLMs, unsupervised clustering, regression & classification on health-data pipelines.
  
• Work on retrieval pipelines: indexing large metadata-rich documents (free-form text, tables, figures), designing prompting strategies (e.g., ReAct, multimodal-ReAct, thought-prompting) and building reasoning agents that deliver factually accurate, multi-step responses for HCC capture, encounter reconciliation, audit readiness.
  
• Collaborate with cross-functional teams (data engineers, product, domain experts) to integrate models into HDMs risk-adjustment and value-based care platform.
  
• Continuously monitor, evaluate and improve model performance in production: e.g., retrieval accuracy, coherence, factuality, latency, auditability.
  

Preferred candidate profile

• 1-2 years of hands-on experience in Python software development.
  
• Solid fundamentals in data science: statistics, regression/classification, clustering, model evaluation metrics.
  
• Experience with SQL and/or NoSQL databases.
  
• Familiarity with natural-language processing techniques (tokenization, embedding, retrieval, transformer architectures).
  
• Understanding of large language models (LLMs) and retrieval-based pipelines.
  
• Enthusiasm for building end-to-end AI systems (not just prototyping) from ingestion to inference to API.
  
• Strong problem-solving mindset, ability to work in a fast-paced start-up environment, and willingness to learn deeply.

• Experience/configuration in RAG pipelines from complex metadata in documents (e.g., free-form text + tables/figures).
  
• Hands-on with advanced prompting approaches: ReAct (Reason + Act), multimodal-ReAct, thought-prompting, conversational flows (human assistant).
  
• Experience or interest in agentic AI architectures (multi-agent reasoning, chain-of-thought, tool-usage).
  
• Interest or exposure in healthcare domain: risk-adjustment, member-care-gaps, clinical data ingestion, value-based care concepts is a plus.