A hands-on Agent Development Bootcamp from ThinkInAI Community and MemVerge. Build your first deployable AI Agent in real-world business scenarios.
- Practice-Driven, Not Conceptual: Every module is a hands-on project. We skip the abstract theories and focus on what you can build.
- Real & Reusable Outcomes: The agents you build are directly applicable to your business or startup ideas.
- End-to-End Toolchain: Master a complete stack—from LLM orchestration and long-term memory to workflow automation.
| Capability | Value |
|---|---|
| Open WebUIOpen WebUI | Build a local/private AI workstation for a controllable and auditable Agent interface. |
| MemMachineMemMachine | Equip your agent with long-term memory for continuous learning and context tracing. |
| LLM APIs (DeepSeek / doubao / PPIO, etc.) |
Confidently call different models, combining reasoning, generation, and retrieval as needed. |
| Toolchains & Workflow Orchestration | Construct scalable, multi-tool, and multi-stage agent processes. |
Before we begin, please follow the Project Launch Checklist to set up your environment, install necessary software, and configure API keys. This ensures you can start coding from day one.
| Project | Agent Theme | Core Learning Objectives | Practical Highlight |
|---|---|---|---|
Project 1AI Writing |
AI Writing Assistant | How to make a model understand, imitate, and embody your unique writing style. | Build a personalized writing assistant you can deploy and use immediately. |
Project 2Health & Wellness |
AI Health Companion | How to construct a long-term health and wellness agent. | Create a "Family Health Assistant" that can record data, send reminders, and provide care. |
Project 3Learning |
AI Study Partner | How to design a learning companion that evolves and adapts to your study habits. | Own a learning agent that becomes more effective the more you use it. |
| Project | Demo | Highlights |
|---|---|---|
| Project 1 Open-Source Writing Assistant |
 |
Continuously learns from your writing samples to generate content in your voice. |
| Project 2 Health Companion |
 |
Manages family member profiles, health reminders, and multi-turn conversations. |
| Project 3 Learning Partner |
 |
Assigns tasks based on your progress and tracks feedback; supports custom lesson packages. |
- Total Duration: 6 weeks — Complete one deployable agent every 2 weeks.
- Live Workshop (1-2 hours): Instructors break down requirements and demonstrate architecture and key workflows.
- Build Sprint (2 weeks): Self-driven development. We suggest building the core in Week 1 and expanding the toolchain for deployment in Week 2.
- Community Support: Get real-time technical support and project feedback in the dedicated WeChat group throughout the sprint.
- Suggested: 3-4 focused sessions of ~2 hours each.
- Flexibility: All sessions are recorded for flexible learning.
Core instructors from ThinkInAI & the technical team from MemVerge, with deep expertise in AI product, engineering, and operations.
- WeChat Group: Daily Q&A and peer review of projects.
- Resource Library: A curated collection of development scripts, prompt templates, and deployment guides.
Outstanding projects will be featured in our official showcase and may lead to internship or co-creation opportunities with partners.
Upon completion, you will have built and deployed three powerful, reusable agents:
- Project 1 · AI Writing Assistant: A partner that truly understands your brand's voice and can continuously learn from your content.
- Project 2 · AI Health Companion: A family health assistant that records data, provides reminders, and offers support through long-term interactions.
- Project 3 · AI Study Partner: A learning companion that adapts to your pace, assigns tasks, and tracks your progress.
Graduate not just with three production-ready agent templates, but with the skills to adapt this toolchain to any business scenario.
Stop learning. Start building. Deploy your agent. Master your craft.