01
AI & LLM5 courses · 17 modules · capstone: multi-agent research assistant
AI & LLM Engineer
RAG pipelines, tool-calling agents, evals. Ship LLM products that survive contact with production.
open track→
Cohort Q4 2026·self-paced · two AI tracks· enrollment open
$
Self-paced AI training, written like a programmer's notebook - agent-led video, hands-on labs, and courseware you can grep.
LatentRAG trains two careers - AI & LLM Engineering and ML & MLOps - one module at a time. Every module ships five ways: video, markdown, lab, PDF, and runnable code.
0
courses
0
modules shipped
0
months to job-ready
24/7
ai instructor
student.mode — week 6, lab 2
2 tracks · AI/LLM · MLOps
AI tutor online — 24/7
$110K–$380K+ 2026 salary band
scroll
PyTorch·LangChain·Hugging Face·vLLM·MLflow·Airflow·Kubernetes·Docker·Pinecone·CrewAI·PyTorch·LangChain·Hugging Face·vLLM·MLflow·Airflow·Kubernetes·Docker·Pinecone·CrewAI·
// pick your track
Two careers.
02
ML & MLOPS4 courses · 12 modules · capstone: self-healing training pipeline
ML & MLOps Engineer
Docker, Airflow, MLflow, drift monitors. Turn notebooks into pipelines nobody has to babysit at 2am.
open track→
03
FREE TIER6 courses · 20 modules · always free
Shared Foundations
Python, math for ML, Git, data wrangling. The floor both tracks stand on - free forever, no card required.
open track→
// straight from the labs
hybrid_search() — week 6, lab 2.
Every module pairs a lab with runnable sample code. Copy it, break it, ship it — this one teaches dense + sparse retrieval blending in 14 lines.
- → runs in Colab or locally
- → graded lab notebook included
- → PDF + markdown courseware
video markdown lab pdf code
week-06/lab-02_hybrid_search.py
# minimal hybrid retrieval - dense cosine + sparse BM25, blended
import numpy as np
def hybrid_search(query_vec, query, docs, k=5, alpha=0.65):
"""Blend cosine similarity with keyword overlap, then rank."""
dense = np.array([cosine(query_vec, d.vec) for d in docs])
sparse = np.array([bm25(query, d.text) for d in docs])
# min-max normalize both channels before blending
dense = (dense - dense.min()) / (np.ptp(dense) + 1e-9)
sparse = (sparse - sparse.min()) / (np.ptp(sparse) + 1e-9)
scores = alpha * dense + (1 - alpha) * sparse
top = scores.argsort()[::-1][:k]
return [docs[i] for i in top], scores[top]enroll.sh
# one subscription, both career tracks
$ curl -s -X POST https://latentrag.com/api/subscribe \
-d '{"plan":"pro","track":"ai-llm"}'
{"status":"active","modules_unlocked":15,"tutor":"24/7"}// the path
Five phases.
15 courses from first line of Python to production MLOps. Phases 1-2 are shared; you branch at Phase 3.
PH.01
Foundations
Python, math for ML, Git, data wrangling - shared by both tracks.
4 crs
PH.02
Core Machine Learning
Supervised learning, deep learning, APIs and Docker.
2 crs
PH.03
Track Specialization
The branch point - LLM systems or MLOps. Pick one, add the other later.
4 crs
PH.04
Production & Scale
Agents and multi-agent systems, pipelines, serving, drift. Pro tier.
3 crs
PH.05
Capstone & Career
Portfolio capstone, runbooks, interview drills.
2 caps
// start now
Every module ships five ways.
Agent-led video, markdown notes, a graded hands-on lab, printable PDF courseware, and runnable sample code - plus an AI instructor and voice tutor on call, 24/7.
videomarkdownlabpdfcode
✓ welcome aboard. syllabus pdf inbound.
$0 to start
Free tier: all Phase-1 foundations. No card.
5 formats
Video, markdown, lab, PDF, code - every module.
Cancel anytime
$29/mo Pro. Your progress stays yours.