Best B.Tech Final Year Projects for CSE (2026): Ideas, Titles and Guidance

Looking for the best B.Tech final year projects for CSE in 2026? This guide lists 60+ final year project ideas for computer science students, each with a complete project title and a short description.

The ideas cover AI, machine learning, generative AI, web development, cybersecurity, cloud, IoT and blockchain.

We have also added mini project topics for CSE, so you can pick a project you can build, demo and explain in your viva. 

Quick Picker

Find your row by skill level, then jump to that domain below.

• Beginner (basic coding): pick from the web, mobile, phishing detection and fraud detection titles. Examples: Online Examination System with Auto-Grading, Phishing URL Detection using Machine Learning.
• Intermediate (comfortable with one framework): pick from the job portal, plant disease detection, fake news detection and intrusion detection titles. Examples: Skill-Based Job Portal with Resume Parsing, Network Intrusion Detection using Machine Learning.
• Advanced (mentor available): pick from the RAG, LLM evaluation, Kubernetes and blockchain titles. Examples: College FAQ Chatbot using RAG, Microservices Deployment on Kubernetes with Auto-Scaling.

What Makes a Strong CSE Final Year Project?

A strong project solves one problem you can state in a sentence, runs live in front of the examiner, and is something you can explain line by line.

It solves a problem you can name

Write the problem statement in plain terms: who has the problem, what is broken today, and what your system changes. “Students check placement notices on five different channels and miss deadlines” is a problem statement. “A portal for placements” is only a title.

It has a working demo and measurable results

Examiners trust numbers. Report model accuracy and F1 score, page load time, requests per second, or the number of test users. A smaller project that works end to end beats an ambitious one that is unfinished.

You can explain every part

If your project uses a trained model, know what the training data is, how you split it and why you chose that algorithm. If it uses a database, know why you designed that schema. This is where most viva marks are won or lost.

What examiners are tired of seeing

• A generic to-do app, library system or calculator with no extra feature
• “Machine learning” with no evaluation metric
• A demo that only works on one laptop

A basic idea becomes a strong capstone project in computer science when you add roles and permissions, analytics, a recommendation step or a deployment. A job portal with resume parsing and skill matching beats a plain job portal.

How to Choose the Right Project Idea for CSE Final Year

Choose by career goal first, then by your skill level and available time, then by team size. Check your college rules before you start.

Choose by career goal

• Software developer (SDE): choose full-stack web, cloud and DevOps projects. They show APIs, databases, authentication and deployment.
• Data analyst: choose analytics dashboards and forecasting projects. They show you can turn data into decisions.
• ML or AI engineer: choose ML, deep learning and GenAI projects with RAG. They show data preparation, evaluation and deployment.
• Security analyst: choose intrusion detection and vulnerability scanning projects. They show safe, controlled testing skills.

Match the scope to your deadline

If your deadline is close, pick a project with an existing dataset or framework and no hardware. If you have more time, add one extra feature such as analytics or a recommendation step. Choose advanced domains such as deep learning, blockchain or a full LLM application only if you have a mentor to guide you.

Solo or team

In a team of 2 to 4, split the work by module (frontend, backend, model, testing), but make every member able to explain the whole system. Examiners often ask each person about the other modules.

Check your college rules first

Confirm guide approval, the report format and whether your university expects an IEEE-style paper. Rules differ across universities, so ask your guide before you lock a topic.

AI and Machine Learning Project Ideas

AI and ML remain the most requested domain for major project ideas for CSE. These titles use current tools and models, and each one can be demonstrated live with a measurable result.

1. Credit Card Fraud Detection using XGBoost, SMOTE and Explainable AI (SHAP)

Build a model in Python that flags suspicious card transactions. Use SMOTE to handle the rare fraud class, XGBoost for prediction and SHAP to explain each alert. Report precision, recall and F1 score. Level: Beginner.

2. Student Dropout Prediction and Early-Warning Dashboard using Machine Learning and Streamlit

Train a classifier on attendance, marks and engagement data to find students at risk of dropping out. A Streamlit dashboard lists them with the key factors, so teachers can step in early. Level: Beginner.

3. Fake News and Misinformation Detection using BERT and Transformer-Based NLP

Fine-tune a BERT or DistilBERT model on a labelled news dataset to classify articles as real or fake. The web app returns a label with a confidence score and compares results against a simple TF-IDF baseline. Level: Intermediate.

4. Plant Disease Detection from Leaf Images using Transfer Learning (EfficientNet or MobileNetV2) with a Web App

Train a pre-trained image model on the PlantVillage dataset to recognise crop diseases. Farmers upload a leaf photo and the app returns the disease name and confidence. Level: Intermediate.

5. AI-Powered Resume Screening and Job Matching System using NLP and Sentence Transformers

Extract skills from resumes with spaCy and compare them to a job description using sentence embeddings. The system gives a match score and lists missing skills, which reduces manual screening. Level: Intermediate.

6. Face Recognition Attendance System using DeepFace, OpenCV and Anti-Spoofing Liveness Detection

Recognise faces from a webcam to mark attendance automatically. Add liveness detection so a printed photo cannot fool the system, and export daily attendance reports. Level: Intermediate.

7. Real-Time Object Detection and Tracking System using YOLO11 and OpenCV

Detect and track selected objects from a live camera feed using a YOLO model. Measure frames per second and detection accuracy, and show results for different lighting conditions. Level: Advanced.

8. Explainable Loan Approval Prediction System using SHAP and LIME

Predict loan approval from applicant data and explain every decision with SHAP and LIME. This shows fairness and transparency, which examiners and recruiters value in modern ML projects. Level: Intermediate.

9. Chest X-Ray Pneumonia Detection using Convolutional Neural Networks and Grad-CAM Heatmaps

Train a CNN on public chest X-ray images and use Grad-CAM to highlight the region behind each prediction. Present it as a decision-support demo, not a medical diagnosis tool. Level: Advanced.

10. Sign Language Recognition System using MediaPipe and LSTM

Track hand landmarks with MediaPipe and feed the sequences to an LSTM network to recognise signs. The output appears as text in real time, which helps communication for the hearing impaired. Level: Advanced.

11. Driver Drowsiness Detection using MediaPipe Face Mesh and Eye Aspect Ratio

Track eye landmarks from a webcam and calculate the eye aspect ratio. When the eyes stay closed beyond a threshold, the system sounds an alert to prevent drowsy-driving accidents. Level: Intermediate.

12. Deepfake Image Detection using Deep Learning

Train a CNN to separate real faces from manipulated ones using a public deepfake dataset. Evaluate it on unseen images and discuss where it fails. Level: Advanced.

13. Crop Recommendation and Yield Prediction System using Machine Learning and Weather API Data

Combine soil values, rainfall and temperature to recommend a crop and estimate yield. Use Random Forest or gradient boosting and display results in a simple web form. Level: Beginner.

14. Sentiment Analysis Dashboard for Product Reviews using Transformers

Classify reviews as positive, negative or neutral with a pre-trained transformer. A dashboard tracks sentiment over time and highlights the most common complaints. Level: Intermediate.

Generative AI, RAG and LLM Project Ideas

This is the freshest category for 2026 and the best way to stand out, because most students still submit classic classifiers. Do not build a thin chatbot wrapper. Add retrieval, an evaluation step and a failure analysis.

15. College FAQ Chatbot using Retrieval-Augmented Generation (RAG), LangChain and a Vector Database

Index your college prospectus and notices in a vector database. The chatbot retrieves the relevant passages and answers with the source cited, which reduces made-up answers. Level: Intermediate.

16. University Document Q&A Assistant using RAG, FAISS and FastAPI

Students upload syllabus or regulation PDFs and ask questions in plain English. The system splits documents, creates embeddings, retrieves matching passages and shows them beside the answer. Level: Intermediate.

17. Text-to-SQL Natural Language Database Assistant using Large Language Models

A user types a question such as “Which branch had the most placements?” and the system generates a safe SQL query, runs it and shows the result. Add query validation to block harmful commands. Level: Intermediate.

18. AI Study Assistant that Generates Quizzes and Flashcards from Lecture PDFs using LLMs

Extract text from lecture notes and use an LLM to create quizzes and flashcards. Add difficulty levels and track which topics the student answers wrongly. Level: Intermediate.

19. AI Code Review Assistant using an LLM API and GitHub Webhooks

When a pull request is opened, a webhook sends the code diff to an LLM, which posts structured comments on bugs, style and security. Measure how many suggestions are useful. Level: Advanced.

20. Multi-Agent AI Interview Preparation Assistant using LLM Agents and Speech-to-Text

One agent asks questions, another evaluates answers and a third gives feedback. Speech-to-text lets students practise out loud, and the report scores correctness and clarity. Level: Advanced.

21. LLM Evaluation Dashboard to Compare Models on Accuracy, Hallucination Rate, Latency and Cost

Run the same test questions through two or more models and record accuracy, response time and cost. A dashboard compares them, and the numbers form a strong results section. Level: Advanced.

22. Fine-Tuning a Small Open-Source LLM with LoRA for Domain-Specific Customer Support

Fine-tune a small open model on support conversations using LoRA and compare answers before and after training. Document the dataset, settings and limits honestly. Level: Advanced.

23. Multilingual Voice Assistant for Indian Languages using Whisper and an LLM

Convert speech to text with Whisper, generate an answer with an LLM and read it back aloud. Support two or three Indian languages and report accuracy for each. Level: Advanced.

24. Prompt Injection Detection and Guardrails System for LLM Applications

Build a filter that detects prompt injection and unsafe requests before they reach the model. Log every blocked attempt and test the filter against a set of known attack prompts. Level: Advanced.

To avoid the wrapper trap, measure something. For a RAG chatbot, write 30 test questions, record how many it answers correctly with and without retrieval, and show the difference. LLM API usage is allowed in most colleges, but confirm with your guide and state clearly what you built versus what the API does.

Full-Stack and Web Development Project Ideas

Full-stack projects suit students who want software developer roles, because one project shows frontend, backend, database, authentication and deployment skills. Deploy yours and share a live link.

25. Placement Management System with Recruiter Dashboard and Placement Analytics

Create a portal where students apply for drives, recruiters post jobs and the placement cell tracks outcomes. Add role-based login and charts for placements by branch and company. Level: Beginner.

26. Online Examination System with Auto-Grading and Webcam-Based Proctoring

Build timed exams with randomised questions and automatic grading. Add webcam snapshots during the test and a result analytics page for teachers. Level: Intermediate.

27. Skill-Based Job Portal with Resume Parsing and Job Recommendation Engine

Candidates upload resumes, the system extracts skills and recommends matching jobs. Employers get a ranked list of applicants for each opening. Level: Intermediate.

28. Learning Management System with Progress Tracking and Quiz Analytics

Instructors publish courses, assignments and quizzes while students track progress. Analytics show which lessons learners find hardest, and certificates are issued on completion. Level: Intermediate.

29. Multi-Vendor E-Commerce Marketplace with Payment Gateway Integration and Product Recommendations

Sellers list and manage their products, buyers check out through a payment gateway in test mode, and a recommendation step suggests related items. Level: Intermediate.

30. Doctor Appointment Booking Platform with Video Consultation using WebRTC

Patients search doctors, book available slots and join a video call at the appointment time. Handle slot conflicts and send reminders by email or SMS. Level: Intermediate.

31. Real-Time Collaborative Document Editor using WebSockets and Operational Transformation

Multiple users edit the same document at once and see each other’s changes live. Operational transformation or CRDTs keep every copy consistent, which makes this a strong systems project. Level: Advanced.

32. Hostel and Campus Management System with Role-Based Access Control

Manage room allocation, student records, complaints, fees and notices. Different roles such as warden, student and admin see different screens. Level: Beginner.

Mobile App Project Ideas

Mobile projects show that you understand user experience, APIs, notifications and databases. Flutter lets you build for Android and iOS from one codebase.

33. College Bus Tracking App with Live Location and ETA using Flutter and Firebase

The driver’s phone shares its location and students see the bus on a map with an estimated arrival time. Add push notifications when the bus is near. Level: Intermediate.

34. Women Safety App with One-Tap SOS, Live Location Sharing and Fake-Call Feature

One tap sends an alert with the live location to chosen contacts. A fake-call feature and a shake-to-trigger option make it more practical. Level: Intermediate.

35. Personal Finance Tracker with Expense Categorisation and Budget Prediction

Record spending, group it into categories and forecast next month’s expenses with a simple regression model. Show budget alerts and monthly summaries. Level: Beginner.

36. Medicine Reminder App with Prescription Scanning using On-Device OCR (Google ML Kit)

Scan a printed prescription, extract medicine names and set reminders automatically. Add refill alerts and a daily adherence report. Level: Intermediate.

37. Object and Currency Recognition App for the Visually Impaired

Run a lightweight model on the phone to identify objects or currency notes through the camera. The app speaks the result aloud, so it works without a screen. Level: Advanced.

38. Indoor Campus Navigation App using Maps SDK and QR Code Checkpoints

Help new students find classrooms, labs and offices. QR codes at checkpoints give the exact location indoors, where GPS is unreliable. Level: Beginner.

Cybersecurity Project Ideas

Cybersecurity suits students aiming at security analyst roles. Test only on systems you own or have written permission to test, and use a lab setup such as local virtual machines.

39. Phishing URL Detection with a Chrome Browser Extension

Extract features from URLs, train a classifier and connect it to a Chrome extension that warns users before they open a risky link. Report accuracy and false positives. Level: Beginner.

40. Network Intrusion Detection System on the CICIDS2017 Dataset

Train models to separate normal traffic from attacks such as DoS and port scans. Compare algorithms using precision, recall and detection rate. Level: Intermediate.

41. Web Application Vulnerability Scanner using OWASP ZAP API Tested on DVWA

Automate scans against a deliberately vulnerable app and report issues such as XSS and SQL injection with severity ratings. Keep all testing inside your lab. Level: Intermediate.

42. End-to-End Encrypted Chat Application using the Web Crypto API and WebSockets

Generate keys in the browser so only the two chatting users can read messages. Explain key exchange and show that the server stores only encrypted data. Level: Intermediate.

43. Login Anomaly and Account Takeover Detection using Isolation Forest

Learn each user’s normal login pattern by time, device and location, then flag unusual logins for extra verification. Test it on synthetic log data. Level: Intermediate.

44. Zero-Knowledge Password Manager with AES-256 Encryption and a Browser Extension

Encrypt credentials on the client with a key derived from the master password, so the server never sees plain text. Add auto-fill through a browser extension. Level: Advanced.

45. Security Log Analysis and Alerting Dashboard using the ELK Stack

Collect logs with Elasticsearch, Logstash and Kibana, then create dashboards and alert rules for failed logins and suspicious activity. Level: Advanced.

Cloud Computing and DevOps Project Ideas

Cloud projects show that you can deploy, monitor and maintain software, not only write it. Free tiers change, so check each provider’s current terms before you start.

46. CI/CD Pipeline with Automated Testing and Docker Deployment using GitHub Actions

Every code push triggers tests, builds a Docker image and deploys the app automatically. Show a failed test blocking a release to prove the pipeline works. Level: Beginner.

47. Serverless Task Management Application using AWS Lambda, API Gateway and DynamoDB

Build a task manager with no servers to manage. Lambda functions handle requests, DynamoDB stores data and the app scales automatically. Include a cost estimate. Level: Intermediate.

48. Application Performance Monitoring Dashboard using Prometheus and Grafana

Collect metrics such as response time and error rate from a sample app and display them in Grafana. Add an alert rule that fires when latency crosses a limit. Level: Intermediate.

49. Microservices Application Deployment on Kubernetes with Auto-Scaling and Rolling Updates

Split an app into services, containerise them and deploy to Kubernetes. Demonstrate auto-scaling under load and a rolling update with no downtime. Level: Advanced.

50. Multi-Environment Infrastructure Automation using Terraform and Cloud Free Tier

Describe servers, networks and databases as code and create separate test and production environments with one command. Level: Advanced.

Data Science and Analytics Project Ideas

Data projects become stronger when they answer a real question, so explain what the patterns mean instead of stopping at charts.

51. Retail Demand Forecasting using Prophet and LSTM with a Power BI Dashboard

Forecast product demand from historical sales and compare Prophet with an LSTM model using forecast error. A Power BI dashboard presents the results to a store manager. Level: Intermediate.

52. College Placement Analytics Dashboard using SQL and Power BI

Analyse placements by branch, company, package range and year using anonymised data. Insights show which skills and branches see the strongest outcomes. Level: Beginner.

53. Air Quality Index Prediction using Machine Learning and Real-Time API Data

Pull pollutant readings from a public API, train a model to predict the next AQI value and display forecasts for a chosen city. Level: Beginner.

54. Electricity Consumption Forecasting using Time-Series Models and Streamlit

Study usage patterns by hour and season, build a forecasting model and show predicted demand in a Streamlit app with error metrics. Level: Intermediate.

55. Customer Churn Prediction and Retention Insights Dashboard using Machine Learning

Predict which customers may leave and rank the reasons with feature importance. The dashboard suggests retention actions for each customer group. Level: Beginner.

IoT and Edge AI Project Ideas

IoT projects combine software with hardware, so decide early what you will demonstrate live and what you will show in a recorded demo.

56. RFID-Based Smart Attendance System with Web Dashboard using ESP32

Students tap RFID cards, the ESP32 sends the data to a server and a dashboard shows attendance records and reports. Level: Intermediate.

57. Smart Parking System using IoT and MQTT with Real-Time Slot Availability

Sensors detect whether each slot is free and publish updates over MQTT. A web page shows a live map of available spaces. Level: Intermediate.

58. IoT-Based Air Quality Monitoring System with Cloud Dashboard and Threshold Alerts

Read gas and particulate sensors, upload the data to the cloud and show live charts. An alert is sent when readings cross a safe limit. Level: Intermediate.

59. Smart Agriculture System with Soil-Moisture Based Automated Irrigation using IoT and Machine Learning

Sensors track soil moisture and weather, and a model decides when to water. Compare water use against a fixed timer schedule. Level: Advanced.

60. Edge AI Object Detection on Raspberry Pi using TensorFlow Lite

Run an optimised detection model on a Raspberry Pi instead of the cloud. Compare speed and accuracy on the device and discuss the trade-offs. Level: Advanced.

Blockchain Project Ideas

Blockchain projects are advanced, so choose one only if you have a mentor to guide you.

61. Blockchain-Based Academic Certificate Verification System using Ethereum Smart Contracts and IPFS

Store a certificate’s hash on the blockchain and the file on IPFS. Anyone can verify authenticity by uploading the document and checking the hash. Level: Advanced.

62. Supply Chain Tracking System using Hyperledger Fabric

Record each handover of a product on a shared ledger so manufacturers, distributors and retailers see one trusted history. Level: Advanced.

63. Decentralised Voting System using Ethereum Smart Contracts

Voters cast one vote each through a smart contract on a test network. Results are transparent and tamper-resistant, and you can discuss privacy limits. Level: Advanced.

Mini Projects for CSE Students (2nd and 3rd Year)

Mini project topics for CSE are a separate search from final year projects, so if you are in 2nd or 3rd year, start here. A good mini project uses one technology well and is finished completely. The same ideas work as PBL (project-based learning) topics and for IT students.

64. Student Result Management System using HTML, CSS, JavaScript, PHP and MySQL

Store marks in MySQL, let teachers enter results and let students view and download their report cards by roll number.

65. Quiz Application with Live Leaderboard using Flutter and Firebase

Players answer timed questions and see scores update live on a leaderboard stored in Firebase.

66. Weather Forecasting Web Application using JavaScript and a Weather API

Show current conditions and a multi-day forecast for any city by calling a public weather API.

67. Library Management System with QR Code Book Tracking using Python Flask

Generate a QR code for each book and track issue, return and fines through a Flask web app.

68. Spam Email Classifier using Natural Language Processing and Scikit-Learn

Train a Naive Bayes or SVM model on labelled emails and report accuracy on unseen messages.

69. Real-Time Chat Application using Node.js and Socket.IO

Build chat rooms with instant messaging, usernames and online status using WebSockets.

70. Password Strength Analyser and Secure Password Generator using Python

Score passwords on length, variety and common patterns, explain the weaknesses and generate stronger alternatives.

71. Smart Home Controller using ESP32 and a Mobile App

Switch lights or a fan on and off from a mobile app over Wi-Fi, with a status display for each device.

72. Movie Recommendation System using Content-Based and Collaborative Filtering

Recommend movies from genres and similar users’ ratings, and compare both methods on a public dataset.

73. Expense Sharing App for Students with Automatic Settlement

Track shared expenses in a group and calculate who owes whom with the fewest payments.

3 ways to get a project with source code

Option	Best for	Risk	What to check
Free GitHub repositories	Learning the structure	Often broken, outdated or undocumented	Does it run today? Is the setup documented?
Ready-made project from a provider	Tight deadlines	You may not be able to explain the code in the viva	Do you get setup help and a code walkthrough?
Guided build with a mentor	Marks and learning	Needs your regular involvement	A named mentor, regular reviews and your own customisation

How to use any source code well

• Run it end to end on your own machine before you commit to it.
• Change something visible: the dataset, the interface, one feature or the evaluation.
• Read every file once and write a one-line note on what it does.
• Rehearse the demo and the viva with someone who did not build it.

How ECEProjectKart helps

Service	What you get	Best for
Guided build	A mentor, step-by-step reviews, working code and deployment help	Students who want to learn while building
Ready-to-run project	Tested source code, a setup session and a code walkthrough	Students with a tight deadline
Report, PPT and viva prep	Report in your university format, PPT and a mock viva	Students whose code is done and documentation is left

Report and PPT: What to Include

• Problem statement: who is affected, what is broken today, what your system changes
• Objectives: 3 to 5 measurable points
• System architecture and data flow diagrams
• Database schema or model pipeline
• Implementation screenshots
• Testing: test cases, results table, accuracy or performance numbers
• Limitations and future scope

For your computer science project file or report, follow your university’s template. For the PPT, cover the problem, solution, architecture, demo, results, limitations and future work. Add a GitHub README with setup steps and screenshots, and a short recorded demo as a backup.

Viva Questions Examiners Actually Ask

Prepare for these groups of questions and practise answering out loud.

Problem and motivation

• Why did you choose this problem?
• Who will use it and what do they do today without it?

Design choices

• Why this database, framework or algorithm instead of another?
• What alternatives did you consider?

Testing and results

• How did you test it? What was your accuracy or response time?
• What happens when the input is wrong or empty?

Domain questions

• ML: What is overfitting? How did you split train and test data? Why this metric?
• Web: How do you store passwords? How do you prevent SQL injection?
• Cloud: How does your app scale? What does it cost to run?

Sample answers

Question	Model answer
What was your contribution?	“I built the backend API and the database, and I wrote the test cases. My teammate built the frontend, and I can explain how it connects to my API.”
What are the limitations?	“The model was trained on a small dataset, so accuracy drops on unseen images. With more data and augmentation, I would expect it to improve.”
Why did you pick this algorithm?	“I compared three algorithms on the same split. This one gave the best F1 score, and it trains faster.”

Common Mistakes to Avoid

• Picking a topic by its title instead of what you can explain
• Submitting code you have not run, read or understood
• Showing results with no accuracy, speed or test numbers
• A demo that only works on one laptop, with no backup recording
• Leaving the report until the very end

FAQs

Q1. Can I use a ChatGPT or LLM API in my final year project?

Usually yes, but confirm with your guide. State clearly what the API does and what you built: the data pipeline, retrieval, evaluation and interface. Projects that only forward text to an API score poorly.

Q2. How many projects should I show on my resume?

Two or three. Put your final year project first, with a one-line result such as “92% F1 on the fraud dataset” and a live link.

Q3. Is an IEEE-based project required?

It depends on your university. Many colleges accept non-IEEE projects. Ask your guide, and if a paper is required, choose a topic with published baselines to compare against.

Q4. Can a final year project become a research paper?

Yes, if it has a clear contribution and an evaluation against a baseline. Plan for it at the start of the project, not after submission.

Q5. What if my guide rejects my topic?

Ask what exactly the concern was: scope, novelty or feasibility. Return with a narrower version and a one-page outline.

Q6. Can I get a project with source code and a report?

Yes. we provides tested source code, a code walkthrough, a report and PPT in your format, and a mock viva.

Ready to Choose Your Project?

Pick a project that fits your skills and goals, build it with a clear scope, and prepare to defend every part of it. If you want a mentor to help, send ECEProjectKart your branch, semester, university, submission deadline and any topic you are considering. We will help you shortlist, scope and start.

Quick Links:

• AI ML Projects for Final Year Students [2026]
• Python projects for final Year Students [2026]
• Cybersecurity Projects for CSE Final Year
• Deep Learning Projects for Final Year