← All 186 courses
AI-215 AI & Data

Vector Databases & Hybrid Search

Build the retrieval layer every RAG system stands on: choose an embedding model and dimensionality deliberately, pick between HNSW IVF and flat indexes on evidence rather than defaults, and combine dense search with BM25 and a reranker instead of hoping one of them is enough. Learners index the same corpus into pgvector and a dedicated store, tune filtering and fusion weights, and measure recall at k against an exhaustive ground truth instead of eyeballing a handful of queries.

B4Professional
6modules
42skill atoms
3role journeys
Curriculum

What this course covers.

6 modules, 42 named skill atoms. Expand any module to see them.

1Embeddings and Vector Space7 skill atoms
embedding model selection criteriadimensionality vs storage and memory costcosine vs dot product vs L2 distancenormalisation before indexingdomain fit of general-purpose embeddingsMatryoshka truncation trade-offchunk granularity per vector
2Index Types and Trade-offs7 skill atoms
flat exact search as the baselineHNSW M and efConstructionefSearch as the recall knobIVF nlist and nprobe tuningproduct quantisation memory savingsbuild time vs query latencyindex footprint against the RAM budget
3Filtering and Data Modelling7 skill atoms
metadata schema designpre-filter vs post-filter semanticslow-selectivity filters collapsing recalltenant isolation by namespace or collectionpayload storage vs external lookupupsert and soft-delete handlingdocument-to-chunk id mapping
4Hybrid Dense plus Lexical7 skill atoms
BM25 term scoring basicssparse vectors and SPLADEreciprocal rank fusionscore normalisation before weightingexact-match and acronym failures of dense searchquery analysis and stopword handlingalpha tuning on a dev set
5Reranking the Shortlist7 skill atoms
cross-encoder rerankersbge-reranker and Cohere Rerankcandidate depth vs added latencyrerank batch sizingscore thresholds that trigger refusalMMR for result diversitywhen reranking cannot rescue bad retrieval
6Measuring Recall Honestly7 skill atoms
labelled query set constructionrecall at k and MRRnDCG for graded relevanceground truth from exhaustive flat searchp95 latency reported beside recallpgvector vs dedicated store benchmarkregression check on every index rebuild
Where it fits

AI-215 in the role journeys.

This course appears in 3 of our 45 role journeys. Here is what a learner takes immediately before and after it in each.

AI / ML Engineer

Practitioner stage
AI-203AI-215AI-204

GenAI / LLM Engineer

Practitioner stage
AI-307AI-215AI-302

AI Architect

Practitioner stage
AI-310AI-215AI-503

Roles this course serves

The capability ladder

This course is authored to band B4.

Every course we run is written to one rung of the CASI ladder, so a plan can be assembled to take a team from where they are to where they need to be.

What do B1–B6 mean?The CASI Capability Ladder — click to expand

Every course targets a band on the CASI Capability Ladder — our six-band proficiency scale, anchored to open standards (O*NET, ESCO, NICE, NIST AI RMF, Bloom's). A band tells you how deep a course goes, and what evidence proves it.

What the learner can doTypical evidence
B1
AwareUnderstands concepts and vocabulary; uses tools with guidance
Knowledge checks
B2
FoundationPerforms standard tasks correctly in familiar contexts
Guided labs, autograded exercises
B3
PractitionerDelivers complete pieces of work independently
Scenario labs, proctored hands-on exams
B4
ProfessionalHandles production-grade complexity, trade-offs and failure modes
Break-fix drills, design defenses
B5
AdvancedEngineers systems end-to-end under constraints; leads others
Rubric-scored capstones, vivas
B6
ExpertSets direction; recognised authority across teams
Portfolio + panel evaluation

A note on B6. Courses in this catalog target B1–B5. B6 is not taught — it is recognised, through a portfolio and a panel, once someone is setting direction for others. Every journey here is built to land a learner at B5.

Next step

Run AI-215 for your team.

This course runs at several lengths depending on how deep you need to go and how much of it your people already have. Tell us who is being trained and we will scope it.

Add it to a training plan Talk to our team Check your team’s level free