Gradpreneur
Explore Courses
← All courses

Generative AI

AI Generalist Fellowship: Intermediate

Move from using AI tools to building retrieval-powered assistants, agents, automations, and deployed products.

Duration5 classes live-online + 6 months recorded access FormatLive And Recorded LevelIntermediate Next start12 Sep 2026
Register for this CourseView curriculum
AI Generalist Fellowship: Intermediate
Program investment£350Enrolling

Course Overview

Build broad, practical confidence across the AI stack

The Intermediate level starts where confident AI use ends. You move past conversational prompting into structured outputs, retrieval-augmented assistants that answer from your own documents, production-shaped automations, multi-step agents, and AI-powered applications you deploy for real users.

Each of the eight curriculum levels pairs a live class with an assignment that produces something deployable. The fellowship finishes with a shipped AI product presented at demo day. It suits people who can already get good results from an AI assistant and want to build the system rather than operate it.

PrerequisiteComfortable using AI assistants daily; no coding experience required

Curriculum

A clear path from foundations to a finished AI solution

Each level combines practical concepts, guided tool use, and an assignment that moves your skills forward.

Level1

Applied AI Foundations Refresher

A fast, practical reset on models, context windows, tokens, and cost so the rest of the program can move quickly.

What you will learn

  • Compare the leading model families and pick one per task on purpose.
  • Explain context windows, tokens, temperature, and where cost comes from.
  • Audit your current AI use and find the three highest-value gaps.

Build and practise

  • Produce a one-page model-selection guide for your own work.
  • Benchmark one real task across three models and record the trade-offs.
Tools & platformsChatGPTClaudeGeminiOpenRouter
Level2

Advanced Prompting & Structured Outputs

Move from conversational prompting to prompts that return predictable, machine-readable output every time.

What you will learn

  • Apply few-shot, chain-of-thought, and decomposition techniques deliberately.
  • Force reliable JSON and schema-shaped output for downstream systems.
  • Build a small test set and measure whether a prompt change helped.

Build and practise

  • Build a prompt that reliably extracts structured records from messy text.
  • Create a twenty-case test set and score two prompt versions against it.
Tools & platformsChatGPTClaudeJSON schemaGoogle Sheets
Level3

Knowledge Assistants with Retrieval

Give an assistant a controlled knowledge base so its answers come from your documents rather than its training data.

What you will learn

  • Explain embeddings, chunking, and semantic retrieval in practical terms.
  • Design a knowledge base that answers a specific set of questions well.
  • Reduce hallucination with grounding, citation, and refusal instructions.

Build and practise

  • Build a domain assistant that answers from a document collection you own.
  • Test it against twenty real questions and fix the retrieval failures.
Tools & platformsCustom GPTsVector databasesLlamaIndexSupabase
Level4

Generative Media Production Pipelines

Turn one-off image and video generation into a repeatable production pipeline with a consistent visual identity.

What you will learn

  • Hold a consistent style across a full set of generated assets.
  • Use image-to-image editing, inpainting, and reference control.
  • Storyboard, generate, and assemble a short branded video.

Build and practise

  • Produce a ten-asset campaign that holds one visual identity throughout.
  • Deliver a 60-second branded video with AI voice and music.
Tools & platformsMidjourneyComfyUIRunwayElevenLabsSuno
Level5

Workflow Automation with n8n and Make

Build multi-step automations that move real data between real systems, with error handling that survives contact with production.

What you will learn

  • Map a business process into triggers, actions, branches, and decision points.
  • Call APIs, transform payloads, and handle authentication safely.
  • Add retries, alerting, and error handling to an automation you rely on.

Build and practise

  • Build an automation that researches, transforms, and routes information daily.
  • Break it on purpose, then make it recover and notify you when it fails.
Tools & platformsn8nMakeWebhooksREST APIs
Level6

Multi-Step AI Agents

Give a model tools and a goal, then keep it on task across a multi-step job you would otherwise do by hand.

What you will learn

  • Describe how planning, tool use, and memory combine in an agent loop.
  • Define tools with clear inputs, outputs, and failure behaviour.
  • Decide when an agent is the right answer and when a workflow is safer.

Build and practise

  • Build a research or operations agent that completes a real multi-step task.
  • Add a human approval step before any action that cannot be undone.
Tools & platformsn8n AI agentsModel Context ProtocolClaudeChatGPT
Level7

Building AI-Powered Applications

Wrap your model, data, and automation in an interface real users can open, use, and give you feedback on.

What you will learn

  • Scope an AI product around one user problem and one measurable result.
  • Connect an interface to a model, a data source, and an automation.
  • Deploy a working product and collect structured feedback on it.

Build and practise

  • Prototype and deploy an AI-powered application for a chosen audience.
  • Run one structured feedback round and ship the improvements.
Tools & platformsLovableReplitSupabaseVercel
Level8

Capstone: Ship an AI Product

Combine retrieval, automation, agents, and product design into one deployed solution and present it at demo day.

What you will learn

  • Combine prompting, data, automation, and product design in one solution.
  • Deploy and document a portfolio-ready project.
  • Present the problem, approach, result, and next iteration clearly.

Build and practise

  • Ship a deployed AI product and present it at the cohort demo day.
Tools & platformsLearner-selected AI stack

Schedule

Choose your cohort

Published dates and sessions update automatically as the Academy opens new cohorts.

Enrolling

September 2026 Cohort

Ten live Saturday classes from 12 September to 14 November 2026, plus weekly build assignments and tutor review.

Starts
12 Sep 2026
Format
Live And Recorded
Timezone
Europe/London

Published sessions

12 Sep 2026Cohort kickoff & project selection
3 Oct 2026Live build: retrieval-powered assistant
31 Oct 2026Live build: multi-step agent
14 Nov 2026Product demo day

Course Tutors

Learn with active AI practitioners

Matthew Makinde

Matthew Makinde

Course Tutor

Practical guidance that helps learners understand AI and put it to work.

Maro Orode

Maro Orode

Course Tutor

Hands-on support for turning AI ideas into useful products and workflows.

Blessed Oigbochie

Blessed Oigbochie

Course Tutor

Project-led teaching focused on confident, practical use of modern AI tools.

Godwin

Godwin

Course Tutor

Learner-focused instruction across creative AI tools and real-world applications.

Take the next step

Register and secure your place

Complete your details, then continue to Stripe’s secure checkout to pay £350.

✓ Secure payment powered by Stripe✓ Your place is confirmed after successful payment✓ Your details are only used for this registration

Course FAQs

Before you apply

People who already use AI assistants regularly and now want to build with them — analysts, marketers, operators, product people, and founders. You do not need to code, but you should be comfortable with tools and data.