Key Responsibilities
Application Development & Integration:
Build, test, and maintain production-ready GenAI applications using LangChain, LlamaIndex, Python, and microservices architectures.
Implement high-accuracy Retrieval-Augmented Generation (RAG) pipelines for content metadata, sports stats, and knowledge bases.
Integrate LLM endpoints into DirecTV's client applications (Gemini OS, iOS, Android, Smart TV apps) via RESTful APIs and WebSockets.

Model Fine-Tuning & Evaluation:
Fine-tune open-source models (Llama, Mistral) using LoRA/QLoRA techniques for specialized entertainment and customer care domain tasks.
Set up automated evaluation suites (using frameworks like Ragas, TruLens) to measure hallucination rates, latency, relevance, and response accuracy.

Prompt Engineering & Orchestration:
Develop, optimize, and version-control complex prompts and agent workflows.
Implement fallback logic, circuit breakers, and guardrails (NeMo Guardrails, Guardrails AI) to ensure safe, reliable user interactions.

Operations & Collaboration:
Collaborate with MLOps and Cloud Engineers to set up continuous deployment pipelines (CI/CD) for model artifacts and vector index refreshes.
Monitor production inference latency, token usage, and error logs to maintain system reliability.

Qualifications & Requirements
Education: Bachelor's or Master's degree in Computer Science, Data Science, AI, or Software Engineering.
Experience: 3+ years in software engineering with at least 1.5-2 years building and deploying Generative AI or NLP applications in production environments.

Technical Skills:
Strong proficiency in Python and modern API development (FastAPI, Flask).
Hands-on experience with LLM frameworks (LangChain, LlamaIndex), Vector Databases (Pinecone, Chroma, Qdrant), and OpenAI/Bedrock APIs.
Experience with git workflow, Docker, Kubernetes, and cloud deployment on AWS, GCP, or Azure.

Mindset: Strong problem-solving skills, quick adapter to fast-moving AI frameworks, and passion for entertainment and video technologies.