Skip to contentSkip to content
Verified credentials. On-chain. Forever.Learn more
Cover image for Build a Hybrid Search for an Enterprise RAG Knowledge Base
Code

Build a Hybrid Search for an Enterprise RAG Knowledge Base

FreeVerified credential2 weeksIntermediate

Overview

What this challenge is about.

You receive 50,000 internal documents (anonymized policy memos, regulation excerpts, internal FAQs) plus a 300-query benchmark with binary relevance labels for the top-10 returned per query. Implement three retrievers: BM25 baseline (Elasticsearch or rank_bm25), dense (sentence-transformers + FAISS), and a hybrid via reciprocal rank fusion. Evaluate on Recall@10, MRR@10, and nDCG@10 (standard IR metrics). Deliver: indexing + query code, evaluation notebook, and a 4-page memo with a clear recommendation plus per-query-type slices (keyword-heavy vs. natural-language).

CredentialBlockchain-anchored
ShareableLinkedIn-ready
LanguageEnglish
PaceSelf-paced

The Brief

What you'll do, and what you'll demonstrate.

Decide whether hybrid retrieval beats dense-only and pure BM25 for an insurance-internal knowledge base, with evidence per query type.

Earning criteria — what you'll demonstrate

  • Implement BM25, dense, and hybrid retrieval end-to-end
  • Apply reciprocal rank fusion to merge ranked lists
  • Evaluate retrievers with standard IR metrics on labeled relevance
  • Slice evaluation by query type, not just aggregate

Program Fit

Where this fits in your program.

Sharpens the same skills your degree expects you to demonstrate.

Skills

Skills you'll demonstrate.

Each one shows up on your verified credential.

Careers

Roles this prepares you for.

Real titles. Real skill bridges. Pick the one closest to your trajectory.

AI Engineer

Hybrid retrieval is the bread-and-butter of every RAG project; owning the implementation and evaluation is core AI-engineer work at any RAG-adjacent company.

This challenge sharpens

  • hybrid-search
  • bm25
  • faiss

AI Solutions Architect

Defending a retrieval-design choice with sliced evaluation evidence is the AI solutions architect's daily output in enterprise engagements.

This challenge sharpens

  • hybrid-search
  • ir-evaluation
  • reciprocal-rank-fusion

NLP Engineer

Combining classical lexical search with dense embeddings is exactly the toolkit NLP engineers ship in production search systems.

This challenge sharpens

  • bm25
  • dense-retrieval
  • ir-evaluation

One more thing

You can put a credential on your CV by Friday.