GOamplify. 15 Day AI Mastery, Day 15
x2 streak Observer
0 XP
Phase 3, Automation, Agents and Enterprise Systems

Day 15. Enterprise AI, RAG and Monitoring

The model stops being the source of truth and becomes the reader. That is the entire idea.

Day narration
2 Why this matters

RAG splits the job in two. Retrieval finds, generation reads.

A retrieval system finds the passages in your documents relevant to the question, and a generation system answers using only those passages, citing them. That is what makes AI usable on contracts, policies, regulations and client documentation where invention is unacceptable. Tools today: NotebookLM, Expertex.ai, Semrush, Opal.

Mechanically, documents are cut into chunks, each chunk becomes a vector, a numerical representation of its meaning, and your question becomes one too. The system retrieves the chunks closest in meaning, not in keyword, which is why it finds the right passage even when your wording differs entirely from the document’s.

3 Watch first

These teachers did this work publicly. Watch them, then come back.

What is Retrieval-Augmented Generation
by IBM Technology
Learn RAG From Scratch, a Python AI Tutorial from a LangChain Engineer
by freeCodeCamp.org
Production RAG with LangChain and Vector Databases
by freeCodeCamp.org
4 The core

The written lesson. Read it slowly, it saves you later.

NOVA, key ideaRefusal is the feature. An assistant that answers everything is indistinguishable from one that is right.

Two consequences you will meet in practice

Chunking matters: cut too small and a clause loses the context that gives it meaning, cut too large and retrieval gets imprecise. Retrieval failure is invisible: if the right chunk is not retrieved, the model answers confidently from the wrong ones. A citation proves where a claim came from, never that the right source was found.

Motion graphic, documents become searchable meaning
THE CORPUS CHUNKVECTORMATCHretrieved by meaning, not by keyword

Refusal is the feature

A grounded assistant must decline when the documents do not contain the answer. Test this deliberately: ask something your corpus genuinely does not cover, and if it invents, tighten the instruction until it refuses cleanly. Then decide the second refusal class: what it must not answer even when the documents contain the answer. Salaries, disciplinary records, legal positions, anything commercially sensitive. That is a policy decision, and it belongs in writing before deployment.

Source hygiene is a job, not a task

Plant a wrong fact in a source document and your assistant will repeat it, with a citation, forever. Grounded systems inherit the honesty of their corpus exactly. So name an owner: someone responsible for what is in the knowledge base, for removing superseded policies, for versioning, for the update process when a regulation changes. If nobody owns it, the system is accurate on launch day and quietly wrong six months later, which is worse than no system, because people have started trusting it.

Monitoring, and the capstone

Anything running unattended needs a way to be seen. Decide up front what is measured, usage, refusal rate, unanswered questions, response time, cost, what the threshold is and what you do when it is crossed, and who looks, how often, in a calendar, not an intention. The unanswered questions log is the most valuable artefact, it tells you precisely what your knowledge base is missing, in the users’ own words.

The final deliverable is a business case, not a demo, because that is the actual job. Cost, risk, the human in the loop policy from Day 13, the update process, the monitoring plan, and an honest statement of what the system will not do. Anyone can show a working assistant. Very few can show one with an owner, a refusal policy, a monitoring plan and a maintenance process attached. That gap is where the commercial work is in this market.

Motion graphic, the capstone stack
ASSISTANTREFUSALSOWNERMONITORTHE CASEthe demo is the easy part
5 Checkpoint

Three quick questions. Not the exam, just a pulse.

NOVA reacts, nothing is scored, nothing is stored against you.

6 Do the work

Two tracks. Pick yours, produce something.

Student mode

Load your subject notes and set texts into NotebookLM and interrogate them. Ask questions the documents do not answer and confirm it declines instead of inventing. That refusal is the whole point of the technique.

Pro mode

Build the capstone. A grounded knowledge assistant over a real document set, with source citations, a defined refusal behaviour, and a documented update process. Wrap it in the Day 12 automation so the knowledge base refreshes on a schedule. Put monitoring on the public output. Then present it as a fifteen minute business case with cost, risk and the Day 13 human in the loop policy.

7 Exercises and brainstorm

Tick them when they are actually done.

Brainstorm, no ticks, just think
Today's badges
Capstone Architectcomplete Day 15
Three in a rowcheckpoint streak
Day close

Day 15 in progress

0
XP today
0/3
Exercises
no
Deliverable

That is the fifteen days. Next stop: the final examination. NOVA will be in the room, quietly.