July 2023 - Present
May 2022 - June 2023
April 2020 - April 2022
Aug 2023 - Dec 2023
April 2023 - Dec 2023
April 2020 - Present
April 2024 - August 2024
Designed and implemented end to end RAG pipeline from scratch using faiss and llama.cpp.
Created separate library for RAG after deriving llama and embeddings from llama.cpp.
Implemented build system using Conan. CI CD using github action.
August 2023 - March 2024
Implemented API to that obtains sensor data streams and transfers them to machine learning module to deliver inference results back into an alarm sytem.
Implemented lock free ring buffer system that stopped data loss and enabled extremely fast inference. Employed unit tests and CI CD.
Feb 2023 - April 2023
Implementing model for crack segmentation using novel transformer based segmentation architecture inspired from retinal vessel segmentation in medicine literature. Implemented continual learning system based on labelling->training cycle. Obtained high precision of 0.9 and recall of 0.8.
November 2022 - April 2023
Constructed model for accurately reading the analog gauge by determining the position of the dial, minimum and maximum values from the face of the scale after applying
perspective transformation. Minimized expected error rate to 0.15%.
Squeezed the model into smaller size by quantization in C++ using TVM. Deployed it into edge device (NVIDIA jetson nano).
May 2022 - July 2022
Because commonly used algorithms are not suitable for fish eye images, implemented new algorithm that can detect people in such images by including information of angular values of slanted bounding boxes. Implemented labelling to storage in cloud pipeline by connecting CoCo Annotator to AWS S3. Obtained a high mAP score, on par with state of the art YOLO algorithms.
June 2021 - October 2021
Model deployment, system design and maintenance using AWS Amplify, AWS Api Gateway, AWS Lambda, AWS Sagemaker.
Front and backend development using Flutter/Dart/Swift programming languages.
March 2021 – May 2021
Designed MySQL database schema and deployed it in Azure.
Using web application API to connect web app with the database.
Aug 2023 - Dec 2023
Collected and labelled dataset of fax documents containing tables. Implemented novel algorithm to extract its data and obtained 97% accuracy.
Dockerized the model and created http endpoint using torchserve. Deployed the docker container into GCP Artifact registry and created
inference endpoint through Vertex AI endpoint.
April 2023 - Sep 2023
Used controlnet and prompt engineering to create Kanjis from novel fonts.
Creating frontend using Gradio.
Creating fastapi endpoint.
Connecting gradio frontend with fast api
Oct 2023 - Dec 2023
Converting caller's speech into text and using prompt engineering to obtain the best output from Open AI api.
Converting the api response into voice. Create WebRTC bridge between python and node js.
This system is currently being deployed into the company app.
June 2021 - Sep 2021
Designing front end and game logic in Swift programming language.
June 2021 - Sep 2021
Emplyed end to end TDD.
p2p mining; Implementing transfer of currencies to user wallets;
Validation of transactions by miners.
Tohoku University
MSc, Particle Physics [2018 - 2020]
Tohoku University
BSc, Physics [2014 - 2018]
Tokyo University of Foreign Studies
Associate Degree, Japanese [2013 - 2014]
Simulation of particle collisions and their reaction in particle detectors using C++.
Desiging particle detectors using C++ framework.
Analyzing data read by particle detectors using machine learning and deep learning algorithms.
Full Detector Simulation of Pair Monitor and Application of Machine Learning to Determine Determine Beam Size
Nagoya University, Nagoya, Japan [March ’20]
Search for weakly interacting dark matter in the International Linear Collider
University of Texas, Arlington, Texas, USA. [Oct ’18]
English, Business Level, TOEFL iBT 110
Japanese: Business Level
Bengali: Native Level