Agentic AI and Autonomous Workflows
Building AI agents that do work rather than answer questions. Learners design and build multi-step agents with tools, memory and human approval gates, and understand where autonomy is appropriate and where it is dangerous.
Who it is for
Technical professionals, automation engineers, product managers, operations leads and AI champions ready to move from chat to systems that act.
What the course covers
- From assistants to agents: what changes when AI can take actions
- Agent architecture: tools, memory, planning, and the orchestration loop
- Multi-agent systems: specialisation, handoffs and agent-to-agent communication
- Human-in-the-loop design: approval gates, escalation and safe failure
- Building real agents hands-on with current agent frameworks
- Evaluating, monitoring and controlling agent cost and behaviour in production
Frequently asked questions
Who is Agentic AI and Autonomous Workflows for?
Technical professionals, automation engineers, product managers, operations leads and AI champions ready to move from chat to systems that act.
How is the course delivered?
As a live, instructor-led private cohort for your team: 2 day(s), online by default, on site by arrangement, tailored to your industry and tools.
How is it priced?
Per engagement rather than per seat, based on group size, customisation and delivery mode. Tell us your team and goals and you get a written proposal, usually within one business day.
Looking for a certified AI foundation? ICP-FAI, Foundations of AI is the ICAgile-certified companion to every course here. For a phased organisation-wide programme, see the Enterprise AI Readiness Programme, or watch real class feedback from our cohorts.