Final year project ideas you can actually finish.
Most FYPs fail on scope, not on ability. Every idea here is sized for one semester, with a suggested stack and a list of what you would actually need to ship. Pick one, make it yours, and cut it down further if you have to.
- 300 project ideas
Ideas and guidance, not finished projects
These are project ideas and scoping guidance, published free for students to use as a starting point. I do not build, write, or sell final-year projects, and I do not complete coursework for anyone. Take an idea, make it yours, and build it.
300 ideas · page 1 of 13
Multi-Agent Literature Review Assistant with Verified Citations for Students
Students writing a literature review often get chatbot answers full of papers that do not exist. Build a multi-agent assistant that plans a search, queries open scholarly databases, reads abstracts and open-access full texts, drafts a structured review, and refuses to cite anything it cannot resolve to a real DOI or arXiv record with a supporting passage. Then measure whether multiple agents actually beat a single agent on quality and cost.
Tamper-Resistant Dynamic Payment QR Display with Voice Confirmation for Small Shops
Printed payment QR stickers at small shops can be covered with a fraudster's sticker, and fake payment screenshots fool busy shopkeepers. Build a small counter device with a screen that shows a fresh, signed QR for each sale, and a speaker that announces confirmed payments only after the shop's own account receives them. Threat-model the attacks, build the device, and test it against sticker swaps and fake confirmations.
Tenancy Rights Guided Assistant with Rules as Code for Punjab and Lagos Rental Laws
Tenants in Lahore and Lagos often do not know whether a rent increase, an eviction notice or a withheld deposit is lawful, and lawyers are expensive. Build a guided web assistant that asks a tenant a few plain questions and explains, with the exact provisions, what the local tenancy law says about their situation, using rules encoded from the statute and reviewed by a law teacher. It generates a short letter template the tenant can adapt.
AI Incident Log and Postmortem Tracker for LLM Products
When an LLM feature misbehaves (a harmful answer, a leaked system prompt, a wrong refund, a runaway cost spike), most teams handle it in a chat thread and forget it. Build an incident tracker designed for AI failures: capture the exact prompt, context and model version, classify the failure, run a blameless postmortem, turn the case into a regression test, and track whether the fix held.
Privacy-First Legal Aid Intake and Case Routing for NGO Legal Clinics
NGO legal clinics and university law clinics in Pakistan, India, Nigeria and Kenya take requests by phone, walk-in and WhatsApp, and sensitive details about domestic abuse, labour disputes or police cases end up in personal phones and spreadsheets. Build a secure intake system with a guided form, issue classification into a standard legal taxonomy, routing to the right volunteer lawyer, and strict access control, encryption and audit logs. Threat-model it before writing code.
AI API Spend Attribution per Feature and Customer with Budget Guards
AI features have a cost per request that changes with prompt length, model choice and retries, and a single heavy customer or a bug in an agent loop can burn a month's budget in a day. Build a lightweight library and collector that tag every AI call with feature, customer and model, compute cost from token counts, show cost per feature and per customer, detect unusual spend quickly, and enforce budgets before the bill arrives.
Court Cause List and Hearing Date Alerts for Litigants and Small Law Chambers
Missing a hearing date can mean a case is dismissed or delayed for months, yet many litigants and small law chambers still check court cause lists by hand. Build a service that reads publicly published cause lists from a few chosen courts on a schedule, matches them to cases a user is following, and sends reminders the day before. The engineering focus is reliable parsing of inconsistent PDFs and pages, and alerts that never miss a match.
Next 48-Hour PM2.5 Forecasts for South Asian Cities from Open Sensors and Weather Data
Each winter, smog in Lahore, Delhi, Dhaka and Kathmandu closes schools and harms health, and decisions are made with little warning. Build a model that forecasts PM2.5 for the next 48 hours at city sensor locations, using open sensor readings, weather forecasts and fire detections, and compare it against the global CAMS forecast. Present it as a simple outlook a school or clinic can use to plan.
EU AI Act Risk Classification Questionnaire for Software Exporters
Software houses in Pakistan, India and Bangladesh build AI features for European clients, and the EU AI Act can apply to them even though they sit outside the EU. Build a guided questionnaire that walks a developer through the Act's structure (prohibited practices, high-risk categories, transparency duties, general-purpose AI), cites the exact article for every step, and produces a plain-language summary of likely obligations to discuss with a lawyer.
Antenatal Follow-Up Prioritiser: Which Pregnant Women a Health Worker Should Call This Week
A health worker with 150 pregnant women on her list cannot call all of them every week. Build a mobile app and ranking model that suggests who to contact first, based on missed visits, risk factors and engagement history, and measures whether the ranking finds women who would otherwise drop out of antenatal care. The project is about fair, explainable prioritisation under a fixed call budget, not about diagnosing anyone.
TinyML Sound and Vibration Anomaly Detection for Backup Generators
Shops, clinics, mosques and offices rely on diesel or petrol backup generators during power cuts, and they fail at the worst moment. Build a small sensor node that listens to a generator's sound and vibration, learns its normal behaviour on the device itself with a tiny anomaly detection model, and sends an alert when something sounds wrong, without a cloud connection or a data scientist.
Shariah Stock Screening and Dividend Purification Calculator with Transparent Methodology
Muslim investors in Pakistan, the Gulf, Malaysia and the UK want to know whether a listed company meets Shariah screening rules and how much of a dividend to give away as purification, but most tools hide their method. Build a screening web app that reads company financial statements, applies a chosen published screening methodology step by step, shows every ratio and its source line, and calculates purification amounts. Let users compare how different methodologies change the result.
Which LLM Understands Roman Urdu and Bangla Business Messages? A Three-Model Evaluation Harness
Small businesses in Pakistan and Bangladesh get orders and complaints as WhatsApp-style messages in Roman Urdu, Urdu script, Bangla and English mixed together. Build an evaluation harness that compares three LLMs (for example one hosted frontier model, one cheaper hosted model and one open model) on real-shaped tasks such as extracting order details and classifying complaints, with a human-checked gold set, cost per correct answer and a public results page.
Model Card and Datasheet Generator from Training Runs
Model cards and dataset datasheets are the standard way to document what a model does, what data it learned from and where it fails, but they are written by hand at the end, if at all. Build a generator that pulls facts from MLflow runs, dataset files and evaluation results, drafts a model card and datasheet, flags missing or unsupported sections, and keeps the card in sync as the model is retrained.
Pipeline That Turns Provincial Statistical PDFs into a Versioned Open Dataset
Governments publish important statistics as PDF yearbooks and reports: schools, hospitals, crops, prices and population by district. Build a data pipeline that extracts tables from these PDFs, cleans and checks them, tracks every change between editions, and publishes them as tidy, documented, versioned datasets that journalists, students and researchers can query directly.
District Medicine Stockout Early Warning from SMS Stock Reports at Rural Clinics
Rural clinics run out of essential medicines while a clinic 20 kilometres away has boxes about to expire. Build a system where clinic staff report stock on hand by structured SMS or a tiny web form, and a district dashboard forecasts which clinics will stock out in the next weeks and suggests transfers between clinics. The focus is the district view and forecasting from patchy reports, not a single pharmacy's ledger.
Mobile Money Agent Cash and E-Float Forecasting to Prevent Turned-Away Customers
Mobile money agents in Kenya, Pakistan, Bangladesh and Nigeria lose customers when they run out of cash or electronic float, especially on salary days and before holidays. Build forecasting models for each agent's cash and e-float needs from simulated transaction histories with realistic calendars, and a simple app that tells the agent and the aggregator when and how much to rebalance. Measure how many turned-away transactions the forecasts would prevent.
Offline AI Tutor Hotspot on a Raspberry Pi for Schools Without Internet
Many schools have a few phones or tablets but no reliable internet. Build a Raspberry Pi box that runs a small open language model and a curriculum-grounded retrieval system completely offline, creates its own Wi-Fi network, and lets students ask questions about their textbook in Urdu or English from any browser. Teachers control what content is loaded and see which topics students ask about most.
PII and Secret Leak Audit for LLM Chat Logs and Traces with Retention Rules
LLM apps log everything: prompts, retrieved chunks, tool results and answers. Those logs quietly fill up with phone numbers, CNIC numbers, addresses, API keys and other users' data. Build an auditor that scans existing LLM traces, finds personal data and secrets in English, Urdu and Roman Urdu, shows where each leak entered the pipeline, and enforces retention and masking rules going forward.
Government Notification Watcher with Plain-Language Summaries and Topic Alerts
Tax notifications, regulatory circulars and gazette notices change rules for small businesses and citizens overnight, but they are published as scanned PDFs on scattered government websites. Build a monitored pipeline that checks chosen official pages on a schedule, detects new documents, extracts text (with OCR when needed), writes a short plain-language summary with the source linked, and alerts subscribers by topic. Treat it as a reliability and cost engineering project.
Closed-Loop Patient Referral Tracking Between Village Health Workers and District Hospitals
When a village health worker refers a patient to a hospital, nobody usually finds out whether the patient arrived, what happened, or whether follow-up is needed. Build an offline-first referral app and a simple hospital desk screen that close the loop: referral slip with a short code, arrival check-in, outcome note and a follow-up task back to the worker. Measure how many referrals get an outcome recorded.
Glacial Lake Growth Monitoring from Sentinel-2 Imagery for Outburst Flood Risk in Northern Pakistan
Glacial lakes in Gilgit-Baltistan and Chitral are growing as glaciers melt, and some can burst and flood villages downstream. Build a pipeline that maps lake outlines from free Sentinel-2 satellite images every month, tracks area change over several years, flags lakes growing fast, and lists the settlements downstream. Validate the outlines against published inventories and hand digitising.
FOCUS Format Converter for Smaller Cloud Providers and AI API Bills
The FinOps Open Cost and Usage Specification (FOCUS) gives cloud bills a common format, and the big clouds now export it. Startups also pay smaller hosts, AI API providers and SaaS tools whose bills arrive in their own formats. Build converters that turn those bills into valid FOCUS data, a validator, and a dashboard that shows total cost and unit cost (per customer or per feature) across every provider in one place.
Tool Selection Evals and Description Linter for MCP Servers
An MCP server can work perfectly and still fail in practice because models pick the wrong tool or fill its arguments badly. Build a tool that lints MCP tool names, descriptions and schemas against good-practice rules, generates realistic user requests for each tool, measures how often several models choose the right tool with valid arguments, and suggests rewrites that improve the score.
Browse project ideas by domain and difficulty
Each of these is its own page, so you can send someone straight to the AI ideas or the beginner ones.
Stuck on something specific?
Writing only gets you so far. If you want an answer to your situation rather than the general case, book a session and we will work through it together. Sessions are free for approved Sefism members, and a few slots open each week.
- 1:1 Mentorship
- Fresh Graduate Mentorship
- Career Guidance
Brand deals, promos and partnerships: tauseeffayyaz@gmail.com
Follow along
New writing, resources and project ideas land here first.