Pramilix International Private Limited
www.pramilix.in
admin@pramilix.in
2-A/3 S/F,Asaf Ali Road, Turkman Gate Darya Ganj New Delhi
Complete Roadmap: Python AI/ML Developer
PHASE 1: Python Foundations
Goal: Build strong Python fundamentals with problem-solving ability.
Topics:
Syntax, variables, data types
Operators, conditionals, loops
Functions (arguments, scope, recursion)
Data structures (list, tuple, dict, set)
String & list manipulation
OOP (classes, objects, inheritance, polymorphism)
File handling (txt, CSV, JSON)
Exception handling
Modules, __init__, packaging
Assignments:
Pattern programs, recursive problems
OOP mini project: Bank System, Library System
100+ basic problems on LeetCode, HackerRank
Project 1: CLI Contact Manager
CRUD app using file storage (JSON)
PHASE 2: Intermediate Python + Dev Practices
Goal: Learn real-world development with best practices.
Topics:
Lambda, map, filter, reduce
Decorators, generators, context managers
Regex
os, datetime, collections, itertools
Logging
Unit testing (unittest, pytest)
Virtual environments
Git/GitHub basics
Project 2: Expense Tracker App (CLI)
Uses CSV/JSON
Unit tests + GitHub versioned
PHASE 3: Python Web Development with FastAPI
Goal: Build scalable RESTful APIs with production structure.
Topics:
REST concepts (GET, POST, PUT, DELETE)
FastAPI basics: Path, query, body params
Pydantic models, validation
Response models
SQLAlchemy + PostgreSQL
JWT Authentication
Routers, dependency injection
Background tasks, async
Docker basics
Project 3: Blog API with Auth
Features: Login, register, write/edit/delete blog posts, token auth
Project 4: Mini E-Commerce API
Auth, products, cart, checkout, order history
PHASE 4: Database & Data Handling
Goal: Master SQL, NoSQL, data loading, and manipulation.
Topics:
PostgreSQL: Joins, subqueries, indexing
MongoDB basics
Python DB drivers: psycopg2, motor
Pandas: Series, DataFrame, indexing, filtering
Merging, grouping, pivot tables
NumPy: ndarrays, broadcasting, vectorization
Project 5: Retail Sales Data Dashboard (CLI/Notebook)
CSV load + Pandas analysis + simple graphs (matplotlib)
PHASE 5: Core Data Science & Math
Goal: Build a solid base in ML, stats, and DS math.
Topics:
Statistics: mean, median, mode, std dev, percentiles
Probability: Bayes, distributions, combinatorics
Linear algebra: Vectors, matrices, dot product
Calculus basics: Gradient, derivatives (intro only)
Data cleaning, EDA
Visualization: matplotlib, seaborn, plotly
Project 6: Student Performance Predictor
Data cleaning, visualization, linear regression model
PHASE 6: Machine Learning Core
Goal: Master essential supervised & unsupervised ML algorithms.
Topics:
Scikit-learn fundamentals
Train/test splits, pipelines
Regression (linear, ridge, lasso)
Classification (logistic, KNN, SVM, decision trees, random forest)
Cross-validation, grid/random search
Clustering (K-means, DBSCAN)
Dimensionality reduction (PCA, t-SNE)
Project 7: Loan Approval ML App
Train model + FastAPI API for predictions
Project 8: Customer Segmentation (Clustering)
PHASE 7: Deep Learning with TensorFlow & Keras
Goal: Build and train neural networks.
Topics:
Tensors, layers, activation functions
Optimizers, loss functions
Feedforward NN, backpropagation
CNNs: Image classification (MNIST, CIFAR)
RNN/LSTM: Text generation
Transfer Learning (ResNet, VGG)
Model evaluation, tuning
Project 9: Handwritten Digit Recognizer (CNN)
Train & deploy via FastAPI
Project 10: Movie Review Sentiment Analyzer
LSTM + text preprocessing
PHASE 8: Advanced AI, NLP & LLMs
Goal: Work with real-world NLP and LLMs.
Topics:
Tokenization, stemming, lemmatization
TF-IDF, word2vec, embeddings
Transformers: BERT, GPT
Hugging Face Transformers & Datasets
LLM APIs: OpenAI, Ollama, HuggingFace
Prompt engineering, few-shot learning
Project 11: Resume Job Matcher with BERT
Input: Resume + JD
Output: Matching score, top jobs
Project 12: ChatGPT-like Mini Chatbot
LLM (OpenChat, LLaMA) served via FastAPI
PHASE 9: Deployment, DevOps & Portfolio
Goal: Prepare for production deployment and showcase.
Topics:
Docker & Docker Compose
Gunicorn + Uvicorn
CI/CD: GitHub Actions
Cloud: Railway / Render / AWS EC2 / GCP App Engine
Monitor: Prometheus + Grafana
Versioning ML models (joblib, MLflow)
Hosting ML APIs
Final Portfolio:
Resume
GitHub profile
Projects with README, Docker setup, test coverage
LinkedIn Optimization
Mock Interviews
Job Interviews and Placement
