अभ्यास • Week 42

verified_user

Weekly Production Deliverable

Target Output

Deliverable: A working RAG endpoint in Spring AI backed by pgvector returning cited answers. Practiced fresh system-design problem (URL Shortener) complete.

Weeks: W41 W42 Active W43 W44 W45

Monday

0 / 2 done

DSA Weekly contest

1.5h

AI/ML Vector Store Architecture & Embedding Design

Revisit text embeddings through Spring AI's VectorStore abstraction; evaluate embedding dimension size, distance metrics (Cosine vs Euclidean)

2.5h

Tuesday

0 / 2 done

DSA Redo 2 hard problems, weakest topic

1.5h

CORE URL Shortener Scaling & IRCTC High-Concurrency Booking

Caching, database sharding, rate limiting + System Design of IRCTC Railway Booking System (High-concurrency inventory locking, flash-sale traffic handling, queue-based booking pipeline)

2.5h

Wednesday

0 / 2 done

DSA Redo 2 hard problems, second weak topic

1.5h

AI/ML Chunking Strategy & Top-K Experiments

Experiment with chunk size (500 vs 1000 tokens), overlap (100 tokens), and top-k retrieval on your actual corpus

2.5h

Thursday

0 / 2 done

DSA Light mixed practice

1.5h

CORE URL Shortener Complete Design Document

Write up formal design document with component diagram, sequence flows, trade-offs, and failure modes

2.5h

Friday

0 / 2 done

DSA Second weekly contest

1.5h

AI/ML Citation & Grounding Pipeline Design

Design prompt templates that force the model to return citations and source snippets rather than ungrounded assertions

2.5h

⚡ Weekend High-Load Execution • 8.0h Daily Deep Focus Lab

Saturday

0 / 3 done

DSA Review contest mistakes

1.5h

AI/ML Spring AI pgvector Store Integration (4.0h)

Configure PgVectorStore in Spring Boot with Dockerized Postgres; implement document ingestion and embedding generation.

4.0h

JAVA Spring AI Vector Store with pgvector

Spin up PostgreSQL with pgvector in Docker; configure PgVectorStore in Spring Boot; build automated document ingestion pipeline (load → split → embed → store)

2.5h

Sunday

0 / 3 done

DSA Full mixed revision

1.5h

AI/ML Spring AI RAG Pipeline & Advisor Chain (4.0h)

Implement QuestionAnswerAdvisor and custom similarity threshold filters for enterprise RAG in Java.

4.0h

JAVA Spring AI RAG Endpoint Implementation

Build end-to-end RAG endpoint: query vector store for top-k chunks, inject context into prompt, return cited answer; test in Postman with real queries; push to GitHub

2.5h

edit_note

Week 42 Technical Notes & Journal

#LeetCode #Java #DSA #Week42