Capstone projects
8 projects that combine many lessons into something you can show an employer. Each has a business requirement, an architecture, build steps, a Python route and a .NET route, tests and a definition of done. Effort values are study estimates.
-
Capstone 1 Machine learning Intermediate
Customer churn prediction service
A subscription business wants a daily list of customers most likely to cancel, plus a live score on the customer-care screen, so the retention team can call the right people first.
Effort 20–30 h
-
Capstone 2 Machine learning Intermediate
Store demand forecasting
A retail chain wants next-14-day demand per product per store so it orders enough stock without filling the warehouse.
Effort 20–30 h
-
Capstone 3 Machine learning Advanced
Visual defect inspection
A factory wants to flag scratched or dented parts from camera images so inspectors only check suspicious items.
Effort 25–40 h
-
Capstone 4 AI application Intermediate
Company knowledge assistant (RAG)
Employees ask questions about HR and IT policies; the assistant answers only from approved documents, cites the passage and says "I don’t know" when the documents do not cover it.
Effort 20–35 h
-
Capstone 5 AI application Advanced
Order-returns agent with human approval
A shop wants an assistant that can look up orders, check the returns policy and prepare refunds, but a staff member must approve every refund.
Effort 25–40 h · Builds on capstone 4
-
Capstone 6 Language models Advanced
Train a small language model on your laptop
Understand language models by building one: train a small transformer on a text collection on your own laptop, then plan how the same code would scale on GPUs.
Effort 25–40 h
-
Capstone 7 Language models Advanced
Private assistant running fully on a laptop
A law office wants a document assistant that never sends text outside the building: a local model summarises and answers questions over local files.
Effort 20–30 h · Builds on capstone 4
-
Capstone 8 Language models Advanced
Fine-tune a small model with LoRA
A support team wants a small model that answers in the team’s own short, polite style and returns the same JSON fields every time. You fine-tune a small open model with LoRA on a few hundred examples and prove with a test set that it beats the plain prompted model.
Effort 20–30 h
No projects match these filters.