Communication Track — Full Curriculum
Active: All 24 weeks · Every session · Non-negotiable
📈 5 Phases of Communication Training
| Phase | Weeks | Focus |
|---|---|---|
| Phase 1 | 1–4 | Technical Vocabulary & Clarity |
| Phase 2 | 5–8 | Structured Explanation (BLUF, Feynman) |
| Phase 3 | 9–12 | Interview Communication (STAR, think-aloud) |
| Phase 4 | 13–16 | Advanced Communication (architecture, senior-level) |
| Phase 5 | 17–24 | Mock Interview Polish |
📚 Phase 1 (Wk 1–4): Technical Vocabulary & Clarity
Goal: Sound like an AI engineer, not a student
Daily habit (5 min):
- Learn 2 new technical terms
- Use them in a sentence out loud
- Explain yesterday's term without notes
Weekly habit:
- Record yourself explaining the week's topic in 2 minutes
- Listen back — count filler words (um, uh, like, you know)
- Reduce filler words by 1 each week
Core vocab to master — Month 1:
LLM, token, context window, temperature, inference, embedding, vector, cosine similarity, system prompt, few-shot, chain-of-thought, RAG, chunking, retrieval, augmentation, tool use, function calling, JSON schema, prompt injection, hallucination, grounding
📝 Phase 2 (Wk 5–8): Structured Explanation
Goal: Explain anything clearly to any audience
The BLUF Method (Bottom Line Up Front)
Bad: "So basically what happens is, you take the text and then it goes through a tokenizer and then the model processes it..."
Good: "The model converts text to tokens, processes them through transformer layers, then generates tokens one at a time."
Techniques to practice:
- 30-second explanation (elevator pitch)
- 2-minute deep dive
- Analogy-first explanation
- Numbers & examples (never vague)
The Feynman Technique (weekly):
- Pick one concept. Explain it as if teaching a complete beginner.
- Where you get stuck = where you don't truly understand.
- Go back and fill that gap.
🎤 Phase 3 (Wk 9–12): Interview Communication
Goal: Sound confident, structured, and senior
The STAR Method
- Situation: Set the scene briefly (1-2 sentences)
- Task: What was your responsibility?
- Action: What did YOU specifically do? (most important)
- Result: Quantify if possible
20 Behavioral Questions to Prepare:
- Tell me about yourself (2-minute pitch)
- Why AI engineering?
- Tell me about your most complex project
- Tell me about a time you failed
- Tell me about a time you disagreed with someone
- Tell me about learning something difficult quickly
- Why Anthropic/OpenAI/Google? (per company)
- Tell me about a decision with incomplete information
- What's your greatest weakness?
- Tell me about a time you influenced without authority
- Describe a project you're most proud of
- Tell me about a time you improved a process
- What if you disagreed with your manager?
- Tell me about handling a tight deadline
- What does good engineering look like to you?
- Tell me about a time you helped a teammate
- Where do you see yourself in 3 years?
- What excites you about AI?
- Tell me about a mistake and what you learned
- Do you have questions for us? (always have 3 ready)
Think-Aloud for Coding:
When solving a coding problem, ALWAYS narrate:
- "First, I'm going to understand the problem — let me restate it..."
- "My initial approach is... but I see a potential issue with..."
- "I'll start with brute force, then optimize..."
- "The time complexity here is O(n) because..."
🏗️ Phase 4 (Wk 13–16): Architecture Communication
Goal: Sound like a senior engineer
- Draw while you talk (never just talk or just draw)
- Name every component explicitly
- State your assumptions first
- Offer tradeoffs proactively ("I chose X over Y because...")
- Ask clarifying questions before designing
Handling "I don't know":
- "I don't know, but here's how I'd find out..."
- "That's outside my experience, but based on similar patterns..."
- "Let me think step by step..." (then actually do it)
🏆 Phase 5 (Wk 17–24): Mock Interview Polish
Full Interview Checklist:
Before the interview:
- Know the company's AI products deeply
- Have 3 questions ready to ask them
- Practice 2-minute "tell me about yourself"
- Every project pitchable in 2 min
During the interview:
- State assumptions before solving
- Think aloud throughout
- Ask clarifying questions
- State time/space complexity
- Test solution with an example
- Offer tradeoffs and alternatives
After the interview:
- Send thank-you email within 24 hours
- Note what to improve
- Research anything you didn't know
📊 Communication Progress Tracker
| Skill | Status | Date Practiced | Notes |
|---|---|---|---|
| BLUF Method | ⭕ Not started | — | — |
| STAR Framework | ⭕ Not started | — | — |
| Feynman Technique | ⭕ Not started | — | — |
| Think-aloud coding | ⭕ Not started | — | — |
| Mock interviews done | 0 | — | — |