Skip to content
Roadmap0/111

Week 5RAG evaluation, privacy, and training-data security

0/6
Active project
Finish P2 core & training-poisoning lab (Days 25–28) $\rightarrow$ Start Project 3 (P3 — Secure Agent & Workflow System, Days 29–30)
Interview focus · 2 h/day outside the 555
Single LLM call vs. workflow vs. agent, bounding agent loops, MCP architecture & tool poisoning, and training-time vs. retrieval-time poisoning.

This week

6 study days
  1. Day 25Add tenant/document ACL tests to the evaluation suite
    S4
  2. Day 26Study PII classification
    S5D4
  3. Day 27Clean Training Baseline
    M2
  4. Day 28Study mitigation options
    M2
  5. Day 29Build a bounded agent loop from scratch with three tools and…
    S3M5
  6. Day 30Study MCP architecture
    S3D2