AI Application Engineering

Five-Day Intensives

• LLM Application Development with Python (35 hours delivered over 5 days, online)

  A 5-day intensive on end-to-end LLM apps in Python: model APIs, prompt pipelines, Pydantic outputs, tool calling, streaming FastAPI, and deployment.

• Building AI Agents with Python and MCP (35 hours delivered over 5 days, online)

  A 5-day intensive on building production AI agents in Python: agent loops, LangGraph and Pydantic AI, MCP servers and clients, orchestration, and evals.

• Mastering Generative AI: From Transformers to Agent Swarms New (35 hours delivered over 5 days, online or on-site)

  A practitioner survey of modern generative AI: frontier and small language models, RAG, the Model Context Protocol, single and multi-agent systems, and evaluation.

• AI and Modern Machine Learning for Software Developers New (35 hours delivered over 5 days, online or on-site)

  The complete machine learning lifecycle in Python: pandas, scikit-learn, XGBoost, and PyTorch, from messy data through trained, evaluated models served with FastAPI and ONNX.

• Advanced Machine Learning and Data Engineering New (35 hours delivered over 5 days, online or on-site)

  Production-scale ML on a modern data platform: Spark 4, dbt, Airflow 3, Kafka 4, lakehouse table formats, MLflow 3, distributed training, and automated retraining.

Building AI-Powered Applications

• LLM Application Development with TypeScript and Python (14 hours delivered over 2-3 days, online or on-site)

  End-to-end LLM application development using TypeScript and Python: prompt pipelines, tool use, streaming, structured output, and production deployment.

• Production RAG Systems for Engineering Teams (14 hours delivered over 2-3 days, online or on-site)

  Building production-grade retrieval-augmented generation systems: chunking, hybrid search, reranking, evaluation harnesses, and deployment.

• Designing Multi-Agent Systems (14 hours delivered over 2-3 days, online or on-site)

  Orchestration patterns, planner/worker architectures, agent hand-offs, and failure recovery for production multi-agent systems.

• Voice and Multimodal AI for Developers (14 hours delivered over 2-3 days, online or on-site)

  Real-time voice APIs, vision, document understanding, and multimodal application patterns for software engineers.

Tools, Skills, and MCP

• Model Context Protocol (MCP) for Developers (7 hours delivered over 1-2 days, online or on-site)

  Building and consuming MCP servers to extend coding assistants with internal tools, data sources, and APIs using the Model Context Protocol standard.

• Building Custom Tools and Skills for AI Coding Agents (7 hours delivered over 1-2 days, online or on-site)

  Function calling, tool design, and safe execution sandboxes for extending AI coding agents with custom capabilities.

Quality and Operations

• Evaluating AI Coding Assistants and LLM Apps (7 hours delivered over 1-2 days, online or on-site)

  Eval harnesses, regression suites, and golden datasets for measuring and improving AI coding assistant and LLM application quality.

• LLM Observability and Cost Engineering (7 hours delivered over 1-2 days, online or on-site)

  Tracing, token budgets, caching, and prompt versioning for production LLM applications that are observable and cost-controlled.

• Guardrails for LLM Applications: Safety, Security, and Validation New (14 hours delivered over 2 days, online or on-site)

  Defense-in-depth guardrails for LLM applications: input and output validation, prompt-injection defense, content moderation, and agent safety, mapped to the OWASP LLM Top 10.

• Faster Responses from LLMs: Latency and Throughput Optimization New (14 hours delivered over 2 days, online or on-site)

  Reducing LLM latency and increasing throughput across the application, system, and model layers: streaming, caching, routing, speculative decoding, and quantization.