★ Agentic AI Level 4 · ST1512 100% Levy Funded No Coding

The agentic AI apprenticeship: build agents that do the work.

Prompting was 2024. The 2026 skill is agentic AI: designing, orchestrating and monitoring autonomous agents that plan and execute real work. TESS apprentices learn exactly that on the Level 4 AI & Automation Practitioner apprenticeship (ST1512), with no coding and 100% levy funding.

Ofsted Good provider 72% distinction rate 4.9/5 from 690+ reviews £18,000 funding band
Plain English first

What "agentic AI" actually means.

A chatbot answers a question. An AI agent completes a job: it takes a goal, plans the steps, uses tools and data to execute them, and only comes back to a human when something needs judgement. Agentic AI is the discipline of building, coordinating and supervising those agents so they work reliably inside a real business.

That shift changes what employers need from training. Prompt skills still matter, but the teams creating measurable ROI in 2026 are the ones who can take a process like invoice triage, customer enquiries or report production and hand it to a governed agent, then prove to a manager (or an auditor) that the agent is behaving. That is what this apprenticeship trains people to do.

The curriculum, in four agentic skills

From single agents to orchestrated, monitored fleets.

These four capabilities run through the AI & Automation Practitioner Level 4 curriculum. Every apprentice builds real agents on real work from their own role.

01 · Agent design & build

Design and build AI agents

Scope which workflows genuinely suit an agent, then design and build it: instructions, knowledge, triggers, tool connections and permission boundaries. Built in Copilot Studio, Gemini Gems and AppSheet, with RAG (retrieval-augmented generation) for grounding agents in company knowledge.

02 · Multi-agent orchestration

Orchestrate agents that work together

Coordinate specialist agents into one workflow: supervisor and worker patterns, sequential hand-offs, and routing logic built in Power Automate, Make.com and Zapier. Learners ship multi-agent workflows that pass work between agents and people cleanly.

03 · Autonomous execution

Let agents execute, with guardrails

Move from human-does-everything to agent-does-the-routine: trigger-based autonomous execution with human-in-the-loop checkpoints for the decisions that matter. Error handling, fallbacks and escalation rules are part of every build, not an afterthought.

04 · Agent performance monitoring

Monitor, evaluate and maintain

An agent you cannot measure is an agent you cannot trust. Apprentices learn to maintain the agents they deploy: output evaluation, sampling and spot checks, trust thresholds, audit trails and the maintenance habits that keep automations reliable month after month.

Microsoft Copilot Studio Google Gemini Gems Power Automate Make.com Zapier AppSheet RAG Claude & ChatGPT
Sensible questions

Agentic AI, answered.

What is an agentic AI apprenticeship?

An apprenticeship that teaches people to design, build, orchestrate and monitor AI agents: software that plans and executes multi-step work autonomously rather than answering one prompt at a time. At TESS this is delivered through the Level 4 AI & Automation Practitioner (ST1512), fully funded by the apprenticeship levy.

Does agentic AI need coding?

No. Agents are built and orchestrated in low-code and no-code platforms (Copilot Studio, Gemini Gems, Power Automate, Make.com, Zapier). The orchestration logic is visual, so people in HR, finance, marketing and operations can build production agents without writing code.

What is multi-agent orchestration?

Coordinating several specialist agents to complete one workflow, usually with a supervisor agent routing work to worker agents and escalating edge cases to a human. It is taught and assessed as part of the Level 4 curriculum, and goes deeper on our Build AI Agents and Microsoft Foundry days.

What is agent performance monitoring?

The practice of keeping autonomous agents reliable once deployed: evaluating outputs, sampling decisions, setting trust thresholds, and keeping audit trails your compliance team will accept. ST1512 requires apprentices to maintain the agents they build, not just ship them.

How is this funded?

Through the apprenticeship levy at the £18,000 funding band, so levy-payers pay nothing extra. Non-levy SMEs are typically 95% co-funded. Workshops are self-funded but can often be structured into cohort spend. Check your levy position.

Where do leaders fit in?

Practitioners build the agents; leaders decide where agents are allowed to act. Our AI Leadership Units (AU0009 to AU0011) cover strategy, policy and delivery for the managers governing agentic AI, and the AI & ML Fellowship (L6) is the senior route.

Stop prompting. Start orchestrating.

Tell us which workflows you want agents to take on and we'll map the right route: apprenticeship, workshop, or both. Levy-funded wherever possible.