Projects
Click on the project name to learn more about it
| Name | Type | Period | Description |
|---|---|---|---|
| MemByte | Side | Jan 2023 - Present | Productivity app that helps with taking notes and time-based recall to retain information longer |
| Coin Trail | Side | Dec 2022 - Present | Personal budget and expense tracker that tracks cash flow |
| Armed Vehicle Detection from Micro-UAV | Freelance | May 2022 - Jul 2022 | Worked with Aerospace research team at MIT, Chennai that is building a border surveillance assisted by drones for DRDO |
| Gamegon | Polygon BUILDIT | Nov 2021 - Dec 2021 | A PvP-based gaming platform that uses play-to-earn model |
| PaySol | Solana Building It Out Loud | Aug 2021 - Sep 2021 | Managed a team of 4 to build a QR code-based crypto wallet with built-in messenger |
| StreamrCVAT | Side | Dec 2019 - Aug 2021 | Led a team of 4 that built a video annotation tool that helps to annotate video datasets at 10x speed |
| blockfake.eth | Side | Aug 2019 - Aug 2021 | Fight misinformation using AI and DAO |
| Fiesta | Course work | Jul 2020 - Nov 2020 | Event managment system built using JSP |
| TataDoc | Internship at TCS | Jul 2020 - Nov 2020 | Built a flutter-based mobile chat application that uses Dialogflow to provide online consultation for health-care related queries |
| Online video-based learning | Course work (Extra credits) | May 2019 - Sep 2020 | Visited Bundubera, Jharkhand twice to understand problems related to education and ideate solutions |
| Fashion Intelligence Systems | Flipkart Grid 2.0 (Ideation) | Jul 2020 | Ideated and designed an system architecture that uses ML algorithms to identify fashion trends and recommend products |
| Iris | Course work | Sep 2019 - Mar 2020 | Contributed to Faculty dashboard's backend and DevOps. Led a team of 6 as a scrum master. |
| Vodafone recommender system | VIL Codefest 2020 | Feb 2020 | Built a recommender system using confidential data provided by the company. Shortlisted as top 10 team out of 600. |
| Caduceus | Hackcamp 2020 | Feb 2020 | Build an application that connects a patients to nearby hospitals during COVID, and predicts spread of virus based on affected regions |
| Udacity Nanodegree projects | Udacity | Nov 2019 - Jan 2020 | Completed 5 projects as part of the nanodegree program |
| Early fire detection using image processing techniques | Research | Nov 2018 - Sep 2019 | Used image processing techniques to detect fire and check if it spreads or not |
| American Sign Language detection | IntelAI 2019 | Aug 2019 | Worked on the UI to display model results from the predicted sign language |
| Audio segmentation to separate voice from noise | SLAC 2019 | Jul 2019 | Led a team of 3. Built a LSTM classifier that uses MFCC features to separate out voice from noise of a conversation between ATC and pilot |
| Vehicle plate detection | JU hackathon | Jul 2019 | Led a team of 4. Built an API that integrates OCR model with Android application and displays stats on the webpage |
| Don't overfit II | Kaggle competition | May 2019 | Ranked 621/2317 to build a model with train data containing 250 rows and test it on a 20K rows data while not overfitting the model |
| Automated passenger check-in with indoor navigation at Airports | Airbus FYI (Ideation) | Oct 2018 | Led a team of 4 to ideate and design a solution to speed-up passenger check-in and avoid flight delays. Visited Madurai airport based on special permission from AAI Director |
| Prediction of household electricity consumption | HCL GUVI 2018 | Aug 2018 | Worked on ML models to predict electricity data prediction using regression models |
StreamrCVAT (Dec 2019 - Aug 2021)
📜 Project description
This deeplearning model continuously learns from previous frames' annotation, to annotate the next frame. Using this tool, you can 10x the speed of annotation process.
Gamegon (Nov 2021 - Dec 2021)
📜 Project description
Gamegon allows an individual to play PvP games, where they get a chance to win 2x rewards by staking their desired amount of MATIC tokens.
🏷️ Features
- Don't just play. You play to earn.
- Stake desired amount of tokens to win 2x rewards (only if the player wins)
- Claim rewards instantaneously after you win the game
- Zero transaction fee charged by Gamegon 🎯 Way forward
- Earn a rare NFT as a badge for winning against another player
- Stake tokens and get rewarded at zero gas fees
- Enable swapping between tokens that are to be staked
- More games to be added to Gamegon. (You can find the list of games to be added here)
- Enhance the user-interface
- Add chat feature
- Streaming live games
- Enable staking tokens on the players playing live games
- Build a personalized dashboard to keep track of the games played
- A leaderboard displaying the game wins of players
PaySol (Aug 2021 - Sep 2021)
📜 Project description
PaySol is a chat based non-custodial crypto wallet that empower users to transact SOLs via P2P, keep track of financial records, advertise offerings, earn SPL tokens for every transaction.
🏷️ Features
- Web based non-custodial crypto wallet
- Enables secure and private 1:1 messaging
- Pay by scanning QR code
- Creating accounts KYC-free
- Zero transaction fee charged by PaySol
🎯 Way forward
- Multi-asset management
- Build reward programs for the users
- Build SPL token to redeem the rewards
- Enhancing the Chat UI
- Advertising products/services in the marketplace
- Subscription based billing and invoicing of payments
- Analytics and reports generation
- Building peer-to-peer system for messaging
- Connecting with Serum DEX
Fiesta - Extracurricular activies management system (Jul 2020 - Nov 2020)
📜 Project description
Project done as a part of Net Centric Programming course under Amrita Vishwa Vidyapeetham. Built a J2EE application for event management.
💻 Technology stacks
- Java servlets
- JDBC
- XML, DTD, XSD, and XSLT
- HTML and CSS
TataDoc (Jul 2020 - Nov 2020)
📜 Project description
Built a mobile application using flutter for online consultation for health related queries, supported by intelligent chatbots
A case study on "Online video based learning" (May 2019 - Sep 2020)
📜 Project description
This project was done as a part of Live-In-Labs program under Amrita Vishwa Vidyapeetham.
🎯 Objective
To provide a solution for improving the quality of education in a Government school at Bundubera, Ranchi, Jharkhand.
🛣️ Findings
- The Government school with a total strength of 50+ students had only 5 school teachers. Since the village was remote, the assigned Government school teachers denied teaching them. The student-teach ratio was around 15:1.
- The school did not have proper access to electricity, water, study materials and infrastructure. Moreover, the school lacked safety, since the school did not have fences.
- The Government school students performed low when compared to another private school student of the same age and grade. Based on research, we found that a 4th grade private school student was able to compete equally with a 7th grade Government school student.
- The quality of teaching and study materials were poor.
🧪 Proposed solution
- Raise funding from Amrita Foundation for building a good infrastructure with electricity and provide all other necessary facilities
- Request Government for increase in the teachers count.
- Providing online education containing pre-recorded lectures with animations with an IoT device that requires less or no electricity required.
Iris - Faculty Dashboard (Sep 2019 - Mar 2020)
📜 Project description
Built a Faculty dashboard following certain Software Engineering practices with various tools and technologies. Visit the project URL to know more about the project.
My contributions are
- Worked on the Backend functionalities for the app
- Static code analysis with SonarQube
- Unit testing with Jasmine, Mocha and Chai
- AWS hosting
- Continuous Integration and Deployment using Jenkins and Netlify
Dynamic and Chromatic analysis for fire detection and alarm raising using real-time video analysis (Nov 2018 - Sep 2019)
📜 Project description
Built an image processing algorithm to detect the fire using Image processing and check if fire spreads, and create an early fire alarm raising to prevent damages.
Vehicle Plate Detection - Digitalization (Jul 2019 - Jul 2019)
My team and I built a model for License plate detection, by extraction of the registration number from the License plate. On extraction of the registration number, we will be able to retrieve the details of the particular vehicle with its owner's detail in an Android app made for Traffic police. This project reduces document verification time of the vehicles drastically.
My contribution:
- Built a simple UI
- Routing using Flask
- Connect with Android application
- Database designer and architect
- Deploy in server
Don't Overfit II (May 2019 - May 2019)
Scored an accuracy of 84.9% and was placed 625/2500 participants in this Kaggle competition