Building in public means sharing what you are learning and shipping while you are still building - not waiting until Demo Day for a perfect reveal. Short posts, screenshots, and honest notes about what worked and what broke.

Why this philosophy

• Accountability - a public commit to ship beats a private to-do list.
• Feedback - the cohort, your network, and future employers see progress and can help early.
• Proof of work - in AI roles, a live URL plus a post about how you built it beats a certificate alone.
• Clarity - explaining what you built forces you to understand it (the Feynman test).

Why we do this in the bootcamp

Every session you add a layer to your capstone. Assignments include sharing in Maven or WhatsApp for cohort feedback and posting publicly so you practice the habit hiring managers and clients actually look for: can you ship, explain, and iterate in the open?

You are not performing for likes. You are building a public trail of skills: what you tried, what you shipped, what you learned. That trail compounds - Session 1's post becomes context for Session 5's demo.

How to use it

1. 1

   Share after each session assignment

   When you ship something real (even v0.1), post within 48 hours while it is fresh. Screenshot or short screen recording beats a polished essay.

2. 2

   Name the skill, not just the tool

   Do not only say "I used Cursor." Say what you directed, what you verified, and what broke - that is the engineering or builder skill.

3. 3

   Credit the programme once, then make it yours

   First post can mention The AI Internship / the bootcamp. Later posts focus on your capstone and your domain.

4. 4

   Ask one specific question

   End with one ask: feedback on scope, a UX choice, or a technical tradeoff. Public posts with a question get better replies.

Where to post: LinkedIn, Substack, and share it

Skills to talk about (Engineering track)

Rotate these as you progress - you do not need all of them in Session 1. Name the ones that match what you actually shipped:

• Coding-agent-first development - directing Cursor or Claude Code, reading diffs, verifying output.
• Reliable LLM components - structured outputs, guardrails, token and cost visibility.
• RAG and grounding - retrieval, chunking, answering from your own data.
• Agents and tool use - planning, LangGraph, MCP.
• TRACE evals - measuring quality before you ship.
• Memory and deploy - persistent state, production URLs, observability.