Skip to content
Learn AI
বাংলা

Contents

All contents

Every part and section in reading order. A check mark shows a section you marked as done.

Sections done
0
Total sections
63
Complete
0%
Continue reading →

  1. LLM Fundamentals

    0/10
    1. 01Tokens, tokenization & context cost
    2. 02Model families & trade-offs
    3. 03Closed vs open-weight models
    4. 04Sampling parameters
    5. 05Streaming vs non-streaming
    6. 06Structured outputs & JSON mode
    7. 07Function calling & tool use
    8. 08Multimodal inputs (vision, audio, PDFs)
    9. 09Embedding models & reranker models
    10. 10Reasoning models & extended thinking

  2. Prompt Engineering

    0/8
    1. 01System vs User vs Assistant message design
    2. 02Few-shot, role prompting & personas
    3. 03Chain-of-thought, ReAct & reflection patterns
    4. 04XML / Markdown structuring for reliable parsing
    5. 05Prompt templating & variable injection
    6. 06Prompt versioning — prompts as code
    7. 07Output parsers & graceful failure handling
    8. 08Prompt injection defense at the prompt layer

  3. Context Engineering

    0/7
    1. 01Context window budgeting
    2. 02Conversation summarization & rolling memory
    3. 03Prompt caching strategies
    4. 04Hierarchical memory
    5. 05Context compression & pruning
    6. 06Tool result truncation strategies
    7. 07Lost in the Middle & mitigation

  4. RAG and Knowledge Systems

    0/10
    1. 01Chunking strategies: fixed, semantic, structural, late chunking
    2. 02Vector databases: Qdrant, Pinecone, Weaviate, pgvector, Chroma, Milvus
    3. 03Hybrid search (BM25 + vector) & reciprocal rank fusion
    4. 04Query rewriting, expansion & HyDE
    5. 05Multi-hop retrieval & agentic retrieval
    6. 06Reranking: Cohere, BGE, Voyage
    7. 07Metadata filtering & structured retrieval
    8. 08GraphRAG & knowledge-graph augmentation
    9. 09Document parsing pipelines: Unstructured, LlamaParse, Docling
    10. 10Multimodal RAG: images, tables, charts

  5. Agents and Agentic Systems

    0/10
    1. 01What an agent actually is
    2. 02The four canonical agent loops
    3. 03Single-agent vs multi-agent orchestration
    4. 04State machines and workflow graphs
    5. 05Agent memory: scratchpad, semantic, procedural
    6. 06Human-in-the-loop checkpoints
    7. 07Subagents and delegation patterns
    8. 08Error recovery, retries, and self-correction
    9. 09Long-running agents and durable execution
    10. 10Browser and computer-use agents

  6. Tool Use and Integrations

    0/1
    1. 01Tool use & integrations (incl. MCP)

  7. Inference

    0/1
    1. 01Inference (servers, quantization, providers)

  8. LLMOps and Observability

    0/1
    1. 01LLMOps & observability

  9. Evaluation Engineering

    0/1
    1. 01Evaluation engineering

  10. Cost and Performance Optimization

    0/1
    1. 01Cost & performance optimisation

  11. Safety, Security, and Guardrails

    0/1
    1. 01Safety, security & guardrails

  12. Multimodal Engineering

    0/1
    1. 01Multimodal engineering (vision, voice, video)

  13. AI Application Architecture

    0/1
    1. 01AI application architecture

  14. Visual Learning Resources

    0/10
    1. 01Transformer Explainer: watch a model predict the next token
    2. 02LLM Visualization: a 3D tour inside a GPT
    3. 033Blue1Brown: neural networks and transformers, animated
    4. 04The Illustrated Transformer: Q, K and V, explained with pictures
    5. 05TensorFlow Playground: train a tiny network and break it
    6. 06A Visual Guide to Attention Variants: MHA, GQA, MLA, sliding window and more
    7. 07A Visual Guide to Mixture of Experts: routers, experts and load balancing
    8. 08LLM Inference Handbook: from prompt to first token and beyond
    9. 09Neuronpedia: look at what features inside a model mean
    10. 10Embedding Atlas by Apple: a visual map of your embeddings