RAG Pipelines for Internal Knowledge cover — Chunk, embed, and retrieve internal PDFs without turning your wiki into a black box.

guided builds · Blended

RAG Pipelines for Internal Knowledge

Walk through ingestion, metadata tagging, and retrieval scoring with an emphasis on Thai + English mixed corpora. You will stress-test latency budgets that match real VPN conditions from Bangkhunthian offices to remote islands.

6 weeks · hybrid async Advanced Weekend intensives

14,800 THB

Informational pricing—see Returns & Refunds for eligibility.

Request information

Inside the syllabus

  • Vector store comparison worksheet
  • Chunking lab with overlap math explained plainly
  • Hybrid retrieval exercises
  • Latency tracing with simple dashboards
  • Security checklist for internal URLs
  • Office hours on VPN throttling scenarios
  • Final architecture one-pager reviewed by mentors

Outcomes you can show

  1. Diagram a RAG stack your IT team can defend in a security review
  2. Tune chunk sizes with before/after recall notes
  3. Ship a rollback plan when embeddings drift
Portrait placeholder for Noah Wijaya

Noah Wijaya

Built retrieval systems for healthcare clinics; focuses on bilingual corpora.

Straight FAQ

Hardware expectations?

You need a machine that can run Docker locally with 16GB RAM. Cloud credits are optional, not bundled.

Can we use our own documents?

Yes, if you certify internal sharing permissions; otherwise we provide masked bundles.

Limitations?

We do not cover multimodal ingestion—text and light tabular only in this edition.

Recent notes

VPN latency lab mirrored what our Songkhla office sees—finally a RAG class that names the boring failure modes.

Hannah · Knowledge manager · 4/5 · Google

Hybrid retrieval worksheet is pinned above my desk; still slower than I'd like on Mondays, but the rollback plan section saved a exec readout.

Client in higher-ed