Description:

Build the intelligence layer on top of the knowledge graph for APAS, transforming structured data into deterministic, citation-backed answers and autonomous workflows.
Collaborate closely with the team, ensuring transparent communication and adherence to high standards of accuracy and defensibility.

Requirements:

Minimum 5 years of experience in production ML/AI systems, with at least 3 years focused on LLMs, NLP, or generative AI.
Proficient in Python (3+ years) with frameworks like PyTorch, LangChain, or Hugging Face Transformers.
Deep experience with retrieval-augmented generation (RAG) architectures and hybrid retrieval methods.
Proven ability to build agentic systems involving multi-step reasoning and task orchestration.
Strong skills in prompt engineering and techniques for hallucination control and citation accuracy.
Experience integrating LLM applications with knowledge graphs and structured data sources.
Familiarity with function-calling and tool-use models, both open-source and proprietary.
Experience fine-tuning LLMs for specific applications and designing evaluation frameworks for output quality.
Comfortable with cloud infrastructure (AWS or Azure) and collaborating with DevOps.
Strong communication skills to explain model behavior to non-technical stakeholders.

Skills