RAG: AI That Knows Your Documents
Answer from a business's own files instead of the whole internet.
- 5 lessons
- about 10 minutes (estimate)
- Taught by Ali
- Free
- Video in اردو
- Exercises in اردو
- 10 quiz questions
- 5 exercises
That was a one-minute preview. The full course video is for signed-in learners. Sign in to watch it here, or create a free BvLogic ID; it takes a minute, and this course is added to it.
What you'll be able to do
- Understand why a model needs your documents.
- Understand the pieces well enough to build one.
- Make the decisions that determine quality.
- Measure it rather than trusting a demonstration.
- Deploy it responsibly.
Course content
- 1 The problem this solves Understand why a model needs your documents.
- 2 How it works Understand the pieces well enough to build one. Sign in to open
- 3 Building one that works Make the decisions that determine quality. Sign in to open
- 4 Testing whether it is any good Measure it rather than trusting a demonstration. Sign in to open
- 5 Running it in production Deploy it responsibly. Sign in to open
Read what each lesson covers
- The problem this solves. A model has never seen your documents. Retrieval finds the relevant passages first, so it answers from current material and can cite where the answer came from.
- How it works. Chunking, embedding, storage and search, then generation, with a decision at each stage, starting with a chunk size of a few hundred words and some overlap.
- Building one that works. Clean and de-duplicate the documents first, extract text properly, keep source and date with each chunk, chunk by structure, and retrieve more before narrowing.
- Testing whether it is any good. Thirty to a hundred real questions with known answers. Measure retrieval separately from answers, and count confident wrong answers on their own.
- Running it in production. Filter by permission at retrieval time, keep the index current, log the questions, let people report wrong answers, and watch the cost.
Who this course is for
- Developers asked to build a chatbot that answers from company documents.
- Managers weighing an internal assistant for policies, manuals or product documentation.
- Anyone who has seen a chatbot confidently answer about 'a generic company' instead of theirs.
Try the first exercise (in Urdu)
کوئی ایک سوال چنیں جو عملہ بار بار پوچھتا ہے۔ دیکھیں اس کا جواب آپ کی دستاویزات میں کہاں ہے۔ اگر تین جگہ ہے اور مختلف ہے تو پہلا مسئلہ وہی ہے۔
Using this in Pakistan
- AI automation entry roles advertised in September 2026 (Junior AI Automation Engineer, AI Automation Specialist, AI Video Content Creator) are mostly in Lahore; one Lahore junior role offered Rs 50,000 to 80,000 a month. Source
- The data is still thin: on 30 September 2026 Rozee.pk returned about 110 'AI automation' results across Pakistan and 1 in Faisalabad. Source
- The tools these adverts name: n8n, Make, Zapier, OpenAI and webhooks. Source
Questions
Is it really free?
Yes. No payment and no card; lesson 1 needs no account.
Is there a certificate?
Yes, when you complete every lesson.
Can I learn in Urdu?
Every lesson has an Urdu exercise alongside the English reading.
How long does it take?
About 10 minutes of reading, plus the exercises.
What comes next?
The paid Applied LLM Engineering programme, in Faisalabad and live online.
Applied LLM Engineering
Applied LLM Engineering
This free course is the introduction. The programme is taught by a trainer, marked by the academy, and carries the syllabus in full: the part an employer asks about.
See the full syllabus and fee