Enterprise Agentic AI · Bootcamp

Enterprise AI System Design.

Learn to take an agentic AI application from architecture through production readiness. Enterprise architects, CTOs, and practitioners with enterprise AI experience guide the work through a continuous build and capstone.

Explore the draft syllabus ↓Ask about the course

Full draft syllabus · six weeks

Learn the ideas, then build them into a system.

Each week combines technical foundations with implementation work on one evolving AI application. Cohort dates and final course design are still being developed.

Week 01

Agentic system design

Theory

  • Compare single-agent and multi-agent architectures.
  • Engineer the context agents need to work reliably.
  • Plan short-term and long-term memory for agent workflows.
  • Establish initial evaluations for agent behavior.

Coding: Create a scalable, performant agentic application.

Tools

  • Agent design patterns
  • Context management
Week 02

Search and retrieval patterns

Theory

  • Apply document updates without rebuilding the entire vector index.
  • Design background ingestion flows with access controls.
  • Use scalable search approaches to improve the context returned.
  • Build guardrails for both incoming requests and generated responses.

Coding: Design a RAG application that can scale.

Tools

  • Ingestion pipelines
  • Access control
  • Retrieval evaluations
Week 03

Evaluation and agent security

Theory

  • Create evaluation datasets for agent decisions and tool calls.
  • Use an LLM judge and compare its assessments with human review.
  • Automate regression checks before releasing changes.
  • Probe prompt injection and attempts to invoke unauthorized tools.

Coding: Assemble an automated evaluation suite for the application.

Tools

  • Agent evaluations
  • LLM-as-a-judge
  • Regression testing
Week 04

Deployment and CI/CD

Theory

  • Package AI applications and expose them through secure APIs.
  • Connect databases with background workers.
  • Automate test and deployment workflows through CI/CD.
  • Run containerized AI applications in the cloud.

Coding: Deploy the agentic application with secure API access.

Tools

  • Secure APIs
  • Cloud deployment
  • CI/CD
Week 05

Observability, scale, and optimization

Theory

  • Trace requests across agent steps and MCP tool interactions.
  • Track errors, token consumption, and user feedback.
  • Use caching, rate limits, and asynchronous execution to improve performance.
  • Configure alerts to help respond to service outages.

Coding: Add monitoring and performance improvements to the application.

Tools

  • Tracing
  • Monitoring
  • Reliability
Week 06

Production capstone

Theory

  • Review architecture and explain key system design decisions.
  • Assess security, reliability, and readiness for production use.
  • Explore failure scenarios alongside cost and performance tradeoffs.
  • Evaluate the application against realistic use cases.

Coding: Present a production-grade AI application through a live workflow.

Proposed capstone

An enterprise operations knowledge agent.

Build a system around a realistic operations workflow, such as helping a service or plant team investigate an issue using approved internal procedures and records.

01

Grounded investigation

Retrieve relevant guidance and operational context, keep source evidence attached, and identify when information is missing.

ContextRetrieval
02

Controlled next steps

Use permitted tools within clear access boundaries, with a human review point before consequential actions.

ToolsGuardrails
03

Reviewable operation

Demonstrate evaluation results, security checks, deployment flow, traces, and practical failure handling.

EvaluationOperations
Capstone goal

Present a working, production-shaped AI application and explain its design decisions, evidence, limitations, and operating tradeoffs. Final capstone requirements are part of the course design in progress.

Designed for builders and decision makers

Learn with people who have worked on enterprise AI.

The bootcamp is led by enterprise architects, CTOs, and practitioners who bring experience designing and delivering AI systems in enterprise settings.

Architecture

Make system choices explicit

Connect agent patterns, context, retrieval, and controls to the workflow and the needs of the organization.

Engineering

Build as you learn

Develop one application through successive modules, so integration and operational concerns are addressed along the way.

Readiness

Show evidence, explain tradeoffs

Use evaluations, security checks, deployment behavior, and observability to support a capstone review.

Interested in this program? Ask us about the design.

Contact the bootcamp team →