Posted inAI System Design
Day 25: GraphRAG for Call Graphs and Service Maps
GraphRAG for call graphs and service maps — when edges beat flat chunks for microservices reasoning.
CheatCoders
Posted inAI AWS Machine Learning
Day 50: Project: Internal SDK Autocomplete Model
Project: train and gate an internal SDK autocomplete student (Titan/custom) on private SDK data, served behind the router with CI evals.
Posted inAI Machine Learning
Day 24: Query Rewriting and HyDE Without Hallucinated APIs
Query rewriting and HyDE without inventing APIs — expand symbols and aliases; never hallucinate methods.
Posted inAI Machine Learning
Day 49: Eval for Fine-Tunes: Blind A/B on Real Tickets
Ship a fine-tune only when blind A/B on real tickets moves cycle time or defect rate — not when loss curves look pretty.
Posted inAI AWS System Design
Day 23: Metadata Filters and Tenant Isolation
Metadata filters and tenant isolation — KB filters, IAM, and the failure mode of leaking another customer’s snippets.
Posted inAI AWS Machine Learning
Day 22: Rerankers After Cheap Recall
Rerankers after cheap recall — cross-encoders and Bedrock rerank when they buy quality, skip when they blow p95.
Posted inAI Machine Learning
Day 48: License and Training-Data Risk
Know what you can fine-tune on and what you must not scrape — licenses, ToS, and customer data boundaries for training sets.
Posted inAI AWS System Design
Day 47: Serving Custom Models Behind the Same API
Hide Titan vs Claude vs your distilled student behind one versioned API router with shadow traffic and instant rollback.
Posted inAI Machine Learning
Day 21: Hybrid Search: BM25 + Vectors + Symbols
Hybrid search — BM25 + vectors + symbols — fusion ranks that beat either signal alone on code RAG.
Posted inAI Machine Learning
Day 46: Synthetic Data That Doesn’t Clone GitHub Noise
Generate synthetic training data from internal ADRs and failing tests — not scraped GitHub noise that poisons style and license posture.
