nagaraj-biradar

Software engineering

• Async Python and concurrency

  Evidence: Calling LLMs reliably: retries, fallbacks and cost

• System design fundamentals

  Evidence: Reference architecture for enterprise GenAI, AI system design fundamentals

ML foundations

• Transformers and attention

  Evidence: How LLMs work: tokens, context and sampling

• Probability, statistics and linear algebra for ML

  Evidence: Embeddings and vector search

• Embeddings

  Evidence: Embeddings and vector search

LLM engineering

• Tokens, context windows, sampling parameters

  Evidence: How LLMs work: tokens, context and sampling

• Model selection and routing

  Evidence: AI system design fundamentals, Cost and latency engineering, Model selection, routing and gateways

• Caching, batching, streaming

  Evidence: Cost and latency engineering

Production / LLMOps

• Model gateways, fallbacks and rate limits

  Evidence: Calling LLMs reliably: retries, fallbacks and cost, Model selection, routing and gateways

• Cost tracking and unit economics

  Evidence: Calling LLMs reliably: retries, fallbacks and cost, AI system design fundamentals, Cost and latency engineering

• Tracing, logging and metrics

  Evidence: Reference architecture for enterprise GenAI

RAG

• Vector, keyword and hybrid search

  Evidence: Embeddings and vector search

AI security

• Authentication, authorization and tool permissions

  Evidence: Reference architecture for enterprise GenAI