Applied AI & Deep Learning with Python
Learn the mathematics and engineering behind modern machine learning. You start with NumPy, pandas and classical models, move to neural networks in PyTorch, and finish by building retrieval-augmented generation (RAG) pipelines and fine-tuning open-weight language models.
- Duration
- 14 Weeks
- Effort
- 100 Hours
- Curriculum
- 6 Modules · 28 Lessons
- Outcome
- Capstone Project
Why this programme
What you will learn
Mathematical Grounding
Linear algebra, probability and optimisation explained through code.
Deep Learning in PyTorch
Build, train and evaluate neural networks, from MLPs to transformers.
Applied LLM Systems
Embeddings, vector search, RAG pipelines and parameter-efficient fine-tuning.
Learning outcomes
By the end of this course you will be able to:
- Prepare and analyse data with NumPy and pandas.
- Train and evaluate classical models with scikit-learn and choose appropriate metrics.
- Implement and train neural networks in PyTorch.
- Explain the transformer architecture and attention.
- Build a retrieval-augmented generation pipeline with embeddings and a vector database.
- Fine-tune an open-weight language model with parameter-efficient methods and evaluate it.
Target profiles
Who should enrol
- Python developers who want to move into machine learning engineering.
- Data analysts ready to work with deep learning and language models.
- Engineers integrating LLM features into products.
Career pathways
Machine Learning Engineer • AI Application Developer • Data Scientist
Entry baseline
Course prerequisites
Python programming
Comfortable writing Python functions, classes and working with packages.
Linear algebra basics
Vectors, matrices and matrix multiplication.
Basic statistics
Mean, variance, distributions and probability.
Unsure whether you’re ready? Ask our academic team.
Structured syllabus
Curriculum: 6 modules, 28 lessons and labs
A step-by-step roadmap built around deliberate practice, coding labs and production-grade projects.
NumPy, pandas and the linear algebra, probability and calculus used throughout machine learning.
- NumPy arrays and vectorised computation
- Data wrangling with pandas
- Linear algebra for ML
- Probability, statistics and gradients
- Lab 1.1: Exploratory data analysis on a real dataset
Supervised and unsupervised learning with scikit-learn, feature engineering and model evaluation.
- Regression and classification
- Feature engineering and pipelines
- Model evaluation and cross-validation
- Clustering and dimensionality reduction
- Lab 2.1: Churn prediction model with evaluation report
Tensors, autograd, training loops, regularisation and convolutional networks in PyTorch.
- Tensors and autograd
- Training loops, loss functions and optimisers
- Regularisation and hyperparameter tuning
- Convolutional networks for images
- Lab 3.1: Image classifier trained from scratch
Tokenisation, embeddings, attention and the transformer architecture behind large language models.
- Tokenisation and embeddings
- Attention and the transformer architecture
- Working with pre-trained models
- Lab 4.1: Text classifier using a pre-trained transformer
Vector databases, retrieval-augmented generation, prompt design, evaluation and responsible AI practices.
- Vector databases and similarity search
- Building retrieval-augmented generation pipelines
- Prompt design and structured outputs
- Evaluating LLM applications
- Responsible AI: privacy, bias and safety
- Lab 5.1: Document question-answering assistant with RAG
Parameter-efficient fine-tuning of open-weight models, serving models locally, and a capstone AI application.
- Parameter-efficient fine-tuning (LoRA)
- Serving models locally with Ollama
- Capstone: end-to-end AI application
Capstone project
An end-to-end AI application combining retrieval, a fine-tuned or prompted language model, and an evaluation report.
Upcoming cohorts
Batch schedule & timings
All times are in Indian Standard Time (IST).
Fees and batch dates are sample data pending confirmation
| Batch | Start date | Timings (IST) | Mode | Status | Action |
|---|---|---|---|---|---|
| December 2026 — Weekday EveningBatch DC-AI-2612E | 1 Dec 2026 | 8:00 PM – 10:00 PMTuesday, Thursday & Saturday | Online LiveTen-student batch · ₹7,500/student | Enrolling now | Register for Applied AI & Deep Learning with Python, December 2026 — Weekday Evening |
- Enrolling now
December 2026 — Weekday Evening
Batch DC-AI-2612E
Starts 1 Dec 2026 · Tuesday, Thursday & Saturday, 8:00 PM – 10:00 PM IST · Online Live
Register for Applied AI & Deep Learning with Python, December 2026 — Weekday Evening
Faculty
Taught by practitioners
Instructor profiles for this course will be published before the batch begins.
Frequently asked questions
Contact your batch coordinator as early as possible. Arrangements for catching up on missed sessions are shared at the start of each batch.
You need a computer running Windows 10/11, macOS or Linux with at least 8 GB of RAM (16 GB is recommended for Docker, Kubernetes and machine learning courses), a stable internet connection, and permission to install developer tools such as VS Code, an IDE, Docker Desktop and Git.
Each course page shows its programme fee in Indian Rupees. Online payment is not yet available on this website; once you submit an enrolment request, the DevCrest team will share the confirmed fee and payment options with you.
Details of course completion credentials will be confirmed for each batch. DevCrest certificates are issued by DevCrest and are not a government or university accreditation.
Continue your learning
Ready to join Applied AI & Deep Learning with Python?
Next batch starts 1 Dec 2026. Submit an enrolment request to reserve your place.