Skip to Content
View all events

Creating a Personalized Assistant with Claude Code

Published by O'Reilly Media, Inc.

Intermediate content levelIntermediate

Build a real assistant across tools, channels, and devices

What you’ll learn and how you can apply it

  • Build a personalized assistant in Claude Code for a real workflow
  • Configure memory, tools, skills, and MCP integrations to support repeatable tasks
  • Extend an assistant across devices and external event sources using Remote Control and channels
  • Design subagents and reusable workflows for research, writing, coding, or synthesis tasks

Course description

Claude Code is becoming a practical environment for building assistants that do more than answer one-off prompts. With building blocks such as memory, tools, MCP integrations, browser-connected workflows, reusable skills, subagents, Channels, and Remote Control, it’s now possible to create assistants that support real recurring work across coding, research, writing, note synthesis, and operational tasks.

Apurva Misra shows you how to combine these capabilities to design a personalized assistant for a concrete use case, with an emphasis on workflows, context, and execution. Through live demos and hands-on exercises, you’ll define a use case, choose the right capabilities, build a working assistant in Claude Code, and learn how to make it reachable across interfaces and responsive to outside events.

This live event is for you because...

  • You’re a developer, AI engineer, product-minded builder, or technical operator who wants to move beyond one-off prompting.
  • You work with recurring tasks such as coding, research, writing, note synthesis, documentation, or operational workflows.
  • You want to become more effective at designing assistants that fit how you actually work and can operate across interfaces.

Prerequisites

  • Claude Code installed on your computer and running locally
  • A code editor and terminal
  • Access to a browser
  • Comfort using a terminal, editing files locally, and working with developer tools at a basic level
  • Familiarity with prompting, APIs, or AI tools (useful but not required)

Recommended preparation:

  • Review a short setup guide provided before the course
  • Optional: Enable any Claude Code features needed to experiment with Remote Control or channels

Recommended follow-up:

Schedule

The time frames are only estimates and may vary according to how the class is progressing.

What makes a personalized assistant possible (20 minutes)

  • Presentation: Overview of Claude Code as an assistant environment; memory, tools, skills, MCP, browser-connected workflows, channels, Remote Control, and subagents
  • Group discussion: What makes a personalized assistant different from a chatbot or simple prompt workflow?
  • Q&A

Choosing a real use case and defining assistant behavior (20 minutes)

  • Presentation: How to scope a useful assistant around recurring work
  • Hands-on exercise: Pick a real use case and define the assistant’s job, inputs, outputs, and boundaries
  • Q&A

Core building blocks in practice (40 minutes)

  • Presentation: How to decide what the assistant should remember, fetch, and execute, and when to use MCP or browser-connected workflows
  • Hands-on exercise: Draft the assistant’s memory, tool, and integration strategy
  • Q&A
  • Break

Channels and Remote Control across interfaces (20 minutes)

  • Presentation: How to communicate with an assistant from other interfaces, drive a running session from another device, and react to outside events
  • Demonstration: Compare a Remote Control workflow with a channels-based event flow
  • Q&A

Building the assistant live (50 minutes)

  • Presentation: Step-by-step implementation of a personalized assistant in Claude Code
  • Hands-on exercise: Build or adapt your own assistant along with the demo
  • Q&A
  • Break

Subagents, refinement, and reusable patterns (20 minutes)

  • Presentation: When to use subagents, how to keep assistants maintainable, and how to refine them over time
  • Group discussion: What should be standardized versus customized?
  • Q&A

Wrap-up and Q&A (10 minutes)

  • Presentation: Common pitfalls, next steps, and how to continue improving the assistant

Your Instructor

  • Apurva Misra

    Apurva Misra is a machine learning engineer, AI consultant, speaker, and founder of Sentick. She works with teams building practical AI systems, with a focus on moving from prototypes to reliable production workflows. Her work includes agentic systems, evaluation, context design, and the operational patterns required to make AI systems usable in real organizations.

Skills covered

  • Claude
  • AI Agents
  • C++