Enterprise AI Engineering Advanced

AI Engineering Advanced: Production-Grade Enterprise AI for Senior Teams

Move beyond single agents to scalable, governed multi-agent systems. Teams design planner-executor architectures with memory and tool use, practice semantic routing, and explore Model Context Protocol for controlled, multi-endpoint execution—suitable for regulated, on-prem or air-gapped environments. This course is part of Caltech’s Enterprise AI Engineering series (Foundations, Intermediate, Advanced), a sequence that organizations can target to different skill levels; learners may progress through all three over time. All three tiers belong to the same series and can be sequenced or spaced to cascade learning across roles—enabling cohorts to advance from foundational literacy to applied prototyping and, ultimately, governed multi-agent design.

  • Learners Foundational
  • Time Client definable
  • Duration 5 Days; Definable
  • Program Type Customizable Programs
  • Certificate Type Certificate
  • Format
    Any Format/Location
  • CEUS Available
  • PDUS Available
  • Program Number AIEngA-Custom
  • Fees Group Rate
  • See full course info

Move beyond single agents to scalable, governed multi-agent systems. Teams design planner-executor architectures with memory and tool use, practice semantic routing, and explore Model Context Protocol for controlled, multi-endpoint execution—suitable for regulated, on-prem or air-gapped environments. This course is part of Caltech’s Enterprise AI Engineering series (Foundations, Intermediate, Advanced), a sequence that organizations can target to different skill levels; learners may progress through all three over time. All three tiers belong to the same series and can be sequenced or spaced to cascade learning across roles—enabling cohorts to advance from foundational literacy to applied prototyping and, ultimately, governed multi-agent design.

Print Page
Enterprise AI Engineering Advanced

Program Experience

Participants will architect multi-agent workflows for targeted functions (for example, procurement, HR, or R&D). We implement planning, tool use, and memory; add semantic routing; and integrate MCP for policy-aligned, auditable control across endpoints. The in-class project demonstrates orchestration, governance touchpoints, and scale considerations—evaluated via agent-to-agent reliability tests. Teams leave with reference pipelines, checklists, and a next-step outline to inform internal hardening or subsequent pilots. 

View our instructors

Course Info

Benefits
Topics
Who Should Attend
Schedule
FAQ

Your team will learn to:

  • Architect planner-executor multi-agent systems with tool use and memory

  • Apply semantic routing and policy checks for safe task delegation across agents

  • Use Model Context Protocol (MCP) for multi-endpoint control with auditability

  • Run agent-to-agent reliability tests and design evaluation gates for scale

  • Select deployment options (closed API, open-weight, on-prem/air-gapped) by risk and cost

  • Produce governed, instrumented prototypes suitable for internal review paths

  • Advanced agents: planning, memory, tool use; multi-agent orchestration (planner-executor, semantic routing)

  • MCP for multi-endpoint control; policy hooks and guardrails

  • Reliability: agent-to-agent testing, evaluation scaffolds

  • Practicum options: image captioning, speech-to-speech, adaptive survey agents; functional tracks for procurement, HR, R&D

  • Capstone (education prototype): design a multi-agent system with governance and scale considerations

Senior engineers and technical leads responsible for architecting AI systems at scale, plus managers overseeing governance, risk, and compliance who need to validate multi-agent approaches prior to internal production work or vendor selection.

This customizable 40-hour program can be delivered on-site or at Caltech, virtual or blended. Advanced is often scheduled after a practice interval following Intermediate so teams can consolidate skills before tackling orchestration and scale.

Are multi-agent systems built in this course production-ready? No. Teams construct education-time designs and prototypes only.

Does the course cover specific frameworks and endpoints? Yes. We discuss open- and closed-weight LLMs, LangChain, LlamaIndex, and LangGraph/LangSmith, and we apply MCP for multi-endpoint control where appropriate.

How are governance and reliability addressed in Advanced? We emphasize role-based access, audit trails, evaluation scaffolds, and reliability testing, including agent-to-agent assessment at orchestration scale.

Can this course run behind our firewall and use our internal enterprise model endpoints? Yes. We can deliver the course on-premises, in a VPC, or in air-gapped environments and connect to your enterprise-approved model endpoints (for example, Azure OpenAI or other hosted foundation models, Bedrock in a private VPC, or self-hosted/open-weight deployments). In-class access is used for research and evaluation during the course; no production integrations or changes are performed in class. Any production access, promotion, or change management follows your enterprise policies and remains outside the scope of the course.

How does Advanced fit within the Enterprise AI Engineering series? Advanced is the third tier in a coherent three-course series designed to help organizations cascade learning effectively by targeting each tier to specific skill levels and responsibilities.

Can learners move through the entire series over time? Yes. Many organizations schedule cohorts to progress from Foundations to Intermediate to Advanced as capabilities and needs evolve.

Who owns the outputs created during the course? Your organization owns the outputs produced by participants; Caltech retains ownership of course materials and templates.

Is post-course support available for next steps? Yes. We can host an optional debrief; any ongoing advisory or build work would be covered under a separate agreement.

Our Educators

Our team of educators and guides are experts in their field – engineering pioneers, applied science visionaries, Ted-Talkers, professional facilitators, pilots, problem solvers, marketing mavens, and award-winning authors – who bring academic knowledge, practical approaches, and proven solutions to their programs.

Collectively, they have decades of experience in aerospace, communications, defense, electronics, energy, government, high-tech, pharma/medical devices, and precision manufacturing. 

Photo of Harish Kashyap

Harish Kashyap

Artificial Intelligence, Machine Learning, Data Science

Picture of Mike Frantz, PhD

Mike Frantz, PhD

Artificial Intelligence, Machine Learning, Data Science

Instructors

Picture of Mike Frantz, PhD

Mike Frantz, PhD

Artificial Intelligence, Machine Learning, Data Science

Photo of Harish Kashyap

Harish Kashyap

Artificial Intelligence, Machine Learning, Data Science