No-Code AI Assistants: Custom GPTs, Projects, Gems and Copilot Studio
For non-developers who want to turn a repeatable task into a shared AI assistant: build, test, share and govern one in ChatGPT, Claude, Gemini or Microsoft 365 without writing code.
The curriculum is being prepared and reviewed. Enrollment is not open yet.
About this course
Most teams already have a no-code assistant builder in a tool they pay for. This course shows how each one works and what it is good for: instructions and knowledge files in ChatGPT projects, plus how to audit an existing custom GPT and migrate it to a plugin, since OpenAI has stopped new custom GPT creation and announced their retirement in favor of plugins that bundle skills, apps and reference files; Claude Projects and Skills; Gemini Gems and sharing in Google Workspace; and Microsoft 365 Copilot's Agent Builder alongside Copilot Studio for agents that need connectors, topics and governance.
You will write instructions that hold up, pick and prepare knowledge sources, test an assistant against a small set of real cases, and decide when a no-code build has reached its limit. The final module covers sharing permissions, data exposure through knowledge files and connectors, admin controls, and documenting an assistant so someone else can own it. Product details are stated as of October 2026 and change often; the course teaches you how to check them. The course ends in an automated exam; passing earns a Wirewalk Institute of AI certificate of completion and assessment, not a vendor certification.
What you will be able to do
Choose the right no-code builder (ChatGPT, Claude, Gemini or Microsoft 365 Copilot) for a task based on where the users, data and licenses already are
Write assistant instructions that define role, scope, steps, output format and refusal rules, and revise them from observed failures
Prepare knowledge files and connected sources so the assistant retrieves the right material and does not expose what it should not
Build a working assistant in at least two platforms, including a Copilot Studio or Agent Builder agent grounded in SharePoint content
Test an assistant against a written set of representative and edge-case requests and record pass or fail before sharing it
Migrate an existing custom GPT to a ChatGPT plugin before the published retirement date, listing what does not carry over, such as custom actions, conversations and sharing settings
Share an assistant with the correct audience and document its purpose, sources, owner and review date
Syllabus
6 modules · 17 lessons · about 7 hours
1
Module 1: Assistant Design Before You Click Build
Picking a task worth automating and writing instructions and knowledge that work on any platform.3 lessons
Choosing a Task: Repeatable, Bounded and Checkable
Writing Assistant Instructions: Role, Scope, Steps and Output Format
Knowledge Files: What to Upload, How to Structure It and What to Leave Out
2
Module 2: ChatGPT: Projects, Custom GPTs and the Move to Plugins
Building in ChatGPT projects today, and moving existing custom GPTs to plugins before they retire.3 lessons
ChatGPT Projects: Shared Files, Instructions and Memory
Auditing an Existing Custom GPT: Instructions, Knowledge, Capabilities and Actions
Migrating a Custom GPT to a Plugin: Skills, Apps, Reference Files and What Does Not Carry Over
3
Module 3: Claude Projects and Skills
Project knowledge for context, Skills for repeatable procedures.3 lessons
Claude Projects: Project Knowledge, Instructions and Team Sharing
Claude Skills: Packaging a Procedure in SKILL.md
Projects or Skills: Matching the Feature to the Job
4
Module 4: Gemini Gems in Google Workspace
Building Gems grounded in Drive files and sharing them inside a Workspace domain.2 lessons
Creating a Gem: Instructions, Drive Knowledge and Previewing
Sharing Gems: Viewer and Editor Access and Admin Console Controls
5
Module 5: Microsoft 365 Copilot: Agent Builder and Copilot Studio
Lightweight agents in Copilot Chat versus governed agents with topics, connectors and flows.3 lessons
Agent Builder in Microsoft 365 Copilot: A SharePoint-Grounded Team Agent
Copilot Studio: Topics, Knowledge Sources, Connectors and Actions
Licensing, Copilot Credits and When to Move from Agent Builder to Copilot Studio
6
Module 6: Testing, Sharing and Governing Your Assistant
Proving an assistant works, sharing it safely, and keeping it owned.3 lessons
A Ten-Case Test Sheet: Checking an Assistant Before Anyone Else Uses It
Data Exposure Through Knowledge, Connectors and Over-Broad Sharing
Documenting, Owning and Retiring an Assistant
Exam and certificate
The exam unlocks once you have completed every lesson. It draws 20 questions from a larger bank, spread across the modules, with the answer options shuffled for each attempt. You have 40 minutes and need 80% to pass. After a failed attempt you can try again after 24 hours, up to 3 attempts in total. Feedback is given per module, never as the answer key.
Passing earns a certificate of completion and assessment from the Wirewalk Institute of AI with a unique ID that anyone can check at institute.wirewalk.com/verify. It is not an accredited qualification or a vendor certification. How certification works.
The Wirewalk Institute of AI is independent. This course teaches publicly documented products and is not affiliated with or endorsed by OpenAI, Anthropic, Google, or Microsoft. Product and company names are trademarks of their respective owners.
A hands-on, code-first course on the three major LLM APIs side by side: messages, streaming, structured outputs, tool calling, vision, cost, caching, batch, retries, abstraction layers, cloud routes and testing. You finish by building and hardening a small multi-vendor service.
A practitioner's course on retrieval-augmented generation: when to use it, how to ingest, chunk, embed and index documents, how to combine BM25 and vector search with reranking, and how to measure, debug and operate the result. Every module includes working Python sketches.
A deep dive into the two layers under every semantic search system: embedding models and vector databases. Covers how embeddings are trained and benchmarked, Matryoshka truncation and quantization, ANN index internals (HNSW, IVF-PQ, DiskANN), filtering, multi-tenancy and the operational work of re-embedding and capacity planning.