Danish Shahzad

Personalised learning

AI, Data Science & Python Tutoring

Hands-on, project-based one-to-one lessons in Python, data science, machine learning, deep learning, generative AI, RAG systems, chatbots, LangChain, LangGraph, MCP, automation, and practical AI application development.

Lessons are adapted to the learner's current level, goals, and preferred pace. The focus is on understanding concepts through practical exercises, guided coding, debugging, and real projects rather than only watching lectures or memorising theory.

Extended curriculum

19-stage AI engineering learning roadmap

This roadmap can be adapted into individual lessons or a longer learning plan. Beginners start with the foundations; advanced, infrastructure, and production topics are introduced according to the learner's background and project needs.

01Foundation Python
  • • Python basics: variables, data types, scope, collections, arrays, operators, modules, packages, input, conditionals, loops, functions, args and kwargs, lambdas, strings, files, environments, and external libraries
  • • Object-oriented Python: classes, objects, variables, methods, inner classes, inheritance, polymorphism, encapsulation, abstraction, exceptions, and custom exceptions
  • • Advanced Python: async/await, threading, multiprocessing, decorators, context managers, generators, iterators, metaclasses, design patterns, logging, type hints, testing, dataclasses, TypedDict, protocols, abstract base classes, dependency injection, environment variables, and configuration management
02Data Processing & Visualisation
  • • NumPy arrays, dtypes, reshaping, slicing, broadcasting, vectorisation, filtering, arithmetic, functions, and data persistence
  • • Pandas DataFrames and Series, CSV/Excel/JSON input, missing values, cleaning, GroupBy, joins, merges, window functions, and feature creation
  • • Matplotlib and Seaborn charts, distributions, correlation heatmaps, and data storytelling
03Mathematics & Statistics for AI
  • • Descriptive and inferential statistics; data types; mean, median, mode, variance, standard deviation, correlation, covariance, probability, random variables, and normal distributions
  • • Hypothesis testing with t-tests, chi-square, and ANOVA; Bayesian statistics and maximum-likelihood estimation
  • • Scalars, vectors, matrices, matrix operations, eigenvalues, eigenvectors, derivatives, gradients, optimisation intuition, Adam, and RMSprop
04Machine Learning Foundations
  • • Problem framing; supervised and unsupervised learning; missing values, outliers, scaling, normalisation, encoding, and feature engineering
  • • Train/test splits, cross-validation, evaluation, bias and variance, hyperparameter tuning, regularisation, and model persistence
  • • Linear and logistic regression, decision trees, random forests, SVM, KNN, Naive Bayes, K-Means, XGBoost, LightGBM, CatBoost, PCA, t-SNE, UMAP, forecasting basics, SHAP, and LIME
05Deep Learning
  • • Neurons, neural networks, activation and loss functions, gradient descent, and backpropagation
  • • TensorFlow, PyTorch, and Keras
  • • CNNs, transfer learning, RNNs, LSTMs, GPU use, TensorBoard, and imbalanced datasets
06NLP & Transformers
  • • NLP pipelines, tokenisation, stop words, stemming, lemmatisation, and regular expressions
  • • Bag of Words, TF-IDF, embeddings, and Word2Vec
  • • Transformers, attention, BERT, and the Hugging Face ecosystem
07Foundations of LLMs
  • • Transformer architecture, context windows, tokenisation, sampling, streaming, mixture-of-experts, quantisation, and inference optimisation concepts including vLLM and TensorRT-LLM
  • • Prompt engineering and context engineering
  • • OpenAI, Claude, and Gemini APIs; tool calling, function calling, and structured outputs
08Retrieval-Augmented Generation
  • • RAG architecture, embedding generation, semantic and recursive chunking, multi-query retrieval, hypothetical document embeddings, Self-RAG, corrective RAG, knowledge-graph RAG, offline and online RAG, and agentic RAG
  • • Retrievers, hybrid search, reranking, ColBERT, metadata filtering, and query rewriting
  • • RAGAS, end-to-end evaluation, human evaluation loops, groundedness, and faithfulness
09Vector Databases
  • • Pinecone, FAISS, Qdrant, Weaviate, and Chroma
  • • Indexing, similarity search, dense and sparse hybrid search, HNSW, IVF, upserts, deletes, scaling, benchmarking, backup, and disaster-recovery concepts
10Agentic AI Foundations
  • • Agent definitions and Observe–Plan–Act–Learn, ReAct, agentic search with tools such as Tavily, reflexion, self-critique, Tree of Thoughts, Graph of Thoughts, and Chain of Thought concepts
  • • Short-term, long-term, and semantic memory; planning, research layers, and tool-augmented reasoning
11Agentic Frameworks & SDKs
  • • LangChain, LangGraph, and LangSmith
  • • OpenAI Agents SDK, Claude SDK, agent collaboration, and task decomposition
12Model Context Protocol
  • • MCP architecture, servers, clients, tools, and resources
  • • MCP with APIs, databases, and RAG; cross-platform skills, client and server development, safe authentication, security, and transport concepts
13Spec-Driven & AI-Driven Development
  • • Spec-Driven Development and AI-Driven Development
  • • Spec Kit, SKILL.md, Claude Code, Gemini CLI, and SDK-oriented workflows
  • • General agents and custom agents
14Multi-Agent Systems
  • • Agent orchestration, role-based agents, and supervisor–worker patterns
  • • Collaboration, negotiation, conflict resolution, agent factories, and skill composition
15Evaluation, Safety & Control
  • • Guardrails, policy enforcement, human-in-the-loop oversight, shadow mode, LangSmith datasets, and DeepEval
  • • Agent testing, reliability metrics, and introductory red teaming
16Backend, Deployment & Cloud
  • • FastAPI, Pydantic, and SQLAlchemy ORM
  • • Docker, multi-stage builds, Kubernetes, kubectl-ai, and kagent
  • • AWS, Google Cloud, and Azure deployment concepts
17Dapr & Cloud-Native AI
  • • Dapr state management, Pub/Sub, service invocation, and durable workflows
  • • Cloud-agnostic agents and workflow orchestration
18Observability & MLOps
  • • MLflow, experiment tracking, model versioning, and CI/CD for machine learning
  • • OpenTelemetry, logs, metrics, and traces
19Production & Business Strategy
  • • Digital-FTE and agent-licensing concepts
  • • Cost optimisation, ROI measurement, compliance, data privacy, and AI-system security basics

Modern agentic development

AI-era thinking, agents, and production systems

A complementary track for learners who want to move beyond isolated models and understand how modern AI-assisted work, agent workflows, evaluation, and deployment fit together.

Foundations & AI-Era Thinking

  • • What AI is, prompting, Markdown-to-HTML workflows, code generation, skills and connectors, general agents, and thinking in the AI era
  • • Asking better questions, detecting broken reasoning, systems thinking, first-principles reasoning, clear communication, dilemmas, uncertainty, learning how to learn, and building a thinking portfolio

Agentic Coding & Personal Workflows

  • • Agentic coding, professional cowork workflows, Spec-Driven Development, loop engineering, harness engineering, trusting checkers, and personal agent harnesses
  • • General and coding agents including Claude Code, OpenCode, OpenClaw, and Hermes; effective context engineering; team configuration and CI/CD concepts

From Problems to AI Workers

  • • Recognising agent-suitable problems, problem-solving principles, and moving from one-off tasks to reusable workers
  • • Python in the AI era, connector-native apps, plugins, AI identity, AI-searchable context, digital workers, agent nervous systems, human-agent teams, agent experiences, eval-driven development, deployment, and architecture selection

Agent Applications & Factories

  • • OpenAI Agents SDK, Google ADK, Claude API and Agent SDK, MCP, agent skills, multi-agent reliability, FastAPI, ChatKit, interactive ChatGPT app concepts, vector databases, RAG, SQLModel, augmented memory, TDD, evaluations, knowledge graphs, and GraphRAG
  • • Tutor-style agents, personal AI employees, agent economics, architecture decisions, and publishing concepts

Business-Domain Workflows

  • • Foundational business agents and workflow analysis across finance, legal and compliance, growth, product and value chains, people and efficiency, and innovation
  • • Worker catalogues for sales, finance, and marketing; AI-native company and startup transformation concepts

Production Agent Infrastructure

  • • Docker, Kubernetes, Helm, Kafka, Dapr, ArgoCD, GitOps, observability, cost engineering, traffic engineering, actors, workflows, production security, compliance, disaster recovery, and real cloud-deployment planning

Coverage is tailored to the learner and project. Enterprise architecture, regulated-domain, and cloud-production subjects are taught as concepts and guided practice unless a specific hands-on environment is agreed.

Who can join?

Complete beginners starting Python or artificial intelligence

University students studying programming, data science, machine learning, or AI

Learners who understand theory but struggle to build projects independently

Working professionals who want to apply AI or automation in their work

Developers who want to learn RAG, chatbots, LangChain, or LangGraph

E-commerce professionals interested in analytics and AI applications

Learners who need explanation-led help debugging assignments, portfolio projects, or personal applications

Academic support is explanation-led: I help learners understand, debug, and improve their own work. I do not complete assessed work on a learner's behalf.

What you can learn

Python Programming

  • • Python fundamentals
  • • Variables and data types
  • • Conditions and loops
  • • Functions
  • • Lists, dictionaries, tuples, and sets
  • • File handling
  • • Error handling
  • • Object-oriented programming
  • • Virtual environments
  • • Package management
  • • Debugging
  • • Clean and reusable code

Data Science

  • • NumPy
  • • Pandas
  • • Data cleaning
  • • Missing values
  • • Data transformation
  • • Exploratory data analysis
  • • Feature engineering
  • • Descriptive statistics
  • • Data visualisation
  • • Matplotlib
  • • Seaborn
  • • Plotly
  • • CSV and Excel data
  • • Business data analysis
  • • Inventory and sales analytics

SQL & Analytics

  • • SELECT queries
  • • Filtering
  • • Sorting
  • • GROUP BY
  • • Aggregate functions
  • • Joins
  • • Subqueries
  • • Common table expressions
  • • KPI reporting
  • • Connecting SQL results to Python or dashboards

Machine Learning

  • • Supervised and unsupervised learning
  • • Regression
  • • Classification
  • • Clustering
  • • Train/test splitting
  • • Data preprocessing
  • • Encoding
  • • Feature scaling
  • • Feature selection
  • • Cross-validation
  • • Model comparison
  • • Hyperparameter tuning
  • • Overfitting and underfitting
  • • scikit-learn
  • • XGBoost when relevant
  • • Accuracy, precision, recall, F1-score, confusion matrix, MAE, RMSE, and R²

Deep Learning

  • • Neural network fundamentals
  • • ANN architecture
  • • Activation and loss functions
  • • Optimisers
  • • Training and validation
  • • TensorFlow
  • • Keras
  • • Convolutional neural networks
  • • Introductory image classification
  • • Introductory sequence models
  • • Model evaluation and common training problems

Generative AI

  • • Large language model fundamentals
  • • Prompt engineering
  • • System and user prompts
  • • Structured outputs
  • • OpenAI and Gemini APIs
  • • Token and context concepts
  • • Hallucination reduction
  • • Grounded responses
  • • AI application planning
  • • Responsible use

Retrieval-Augmented Generation

  • • What RAG is and when to use it
  • • Document loading
  • • PDF and text processing
  • • Chunking strategies
  • • Embeddings
  • • Vector search
  • • FAISS
  • • Pinecone
  • • Similarity search
  • • Retrieval pipelines
  • • Source-grounded answers
  • • Citations
  • • Basic RAG evaluation
  • • Common retrieval problems

Chatbot Development

  • • Rule-based versus AI chatbots
  • • LLM chatbot architecture
  • • Conversation history
  • • Prompt design
  • • Retrieval-based chatbots
  • • Customer-support chatbots
  • • Streamlit chat interfaces
  • • FastAPI chatbot backends
  • • Error handling
  • • Deployable project structure

LangChain

  • • Prompt templates
  • • Chains
  • • Document loaders
  • • Text splitters
  • • Embeddings
  • • Vector stores
  • • Retrievers
  • • Retrieval chains
  • • Chat history
  • • Tools
  • • Structured outputs
  • • Practical RAG applications

LangGraph

  • • Graph-based AI workflows
  • • State management
  • • Nodes and edges
  • • Conditional routing
  • • Tool-calling workflows
  • • Memory
  • • Multi-step agents
  • • Human-in-the-loop concepts
  • • Debugging agent workflows
  • • Structured AI applications

Model Context Protocol (MCP)

  • • What MCP is
  • • MCP clients and servers
  • • Tools and resources
  • • Connecting AI applications to external systems
  • • Basic MCP server structure
  • • Tool definitions
  • • Safe permission handling
  • • Testing MCP integrations
  • • Beginner-level examples

FastAPI & Streamlit

  • • Streamlit dashboards
  • • Chat interfaces
  • • Forms and user inputs
  • • Displaying model outputs
  • • FastAPI endpoints
  • • Request and response models
  • • Connecting ML or AI logic to APIs
  • • Project organisation
  • • Basic deployment preparation

Portfolio & Project Guidance

  • • Selecting a realistic project
  • • Defining the problem
  • • Finding or preparing data
  • • Building an end-to-end workflow
  • • Writing a README
  • • Adding screenshots
  • • Explaining project architecture
  • • Preparing GitHub repositories
  • • Project presentations
  • • Understanding and defending technical decisions

How lessons work

01

Goal discussion

We begin by identifying what the learner wants to understand, build, or improve.

02

Level assessment

The lesson is adjusted for beginner, intermediate, or project-focused needs.

03

Live explanation and coding

Concepts are explained through practical examples and guided coding.

04

Hands-on practice

The learner writes and modifies code during the session.

05

Project guidance

Lessons may follow a complete project from idea to implementation.

06

Review and next steps

The session ends with a summary, practice recommendations, and the next learning target.

Lesson formats

Concept ExplanationLive CodingDebugging SupportProject GuidanceCode ReviewPortfolio ReviewInterview Preparation

Session options

Introductory Session

A focused session to discuss your goals, current knowledge, learning challenges, and recommended study path.

Request a Tutoring Session

One-to-One Lesson

A personalised lesson focused on one topic, coding problem, project feature, or technical concept.

Request a Tutoring Session

Project Support Session

Guidance for planning, debugging, improving, or explaining a data science or AI project.

Request a Tutoring Session

Ongoing Learning Plan

A structured sequence of sessions for learners who want regular guidance across Python, data science, machine learning, or AI application development.

Request a Tutoring Session

A practical teaching approach

I currently provide personalised technical guidance to learners through live explanations, coding demonstrations, project walkthroughs, debugging, and structured practice. My approach is based on breaking difficult concepts into smaller steps and helping learners apply them through real code.

I have experience guiding individual learners through Python, data science, machine learning, and AI application projects.

Get in Touch

This form prepares an email in your email application. No form backend or external email provider is currently connected.