Nilambar Halder Tonmoy | Machine Learning Engineer

Hi, I'm Nilambar Halder Tonmoy

A

YOU NEVER FAIL UNTIL YOU STOP TRYING - Albert Einstein

About

I am currently a Ph.D. candidate at Indiana University - Bloomington in Computer Science department. I am actively conducting research in Computer Vision using Machine Learning. Formerly, I was a Software Engineer (Machine Learning) at Graaho, Reston, VA, USA. In their R&D division I mainly researched and worked with Big Data, Data Mining, MLOps, Image & Video Processing. I enjoy researching, learning and implementing various AI solutions. I have significant skills in Tensorflow, Keras, Pytorch, Huggingface Transformers, Django, PostgreSQL, Apache Airflow, Azure ML, AWS, Minio in my three years of professional career. During my undergraduate studies I also had the opportunity to work on Computer Vision, Data Analysis, Supervised-Unsupervised ML Models. I am very enthusiast about working on complex problems and research various technological aspect to derive next-gen solutions.

Languages: Python, Java, JavaScript, Typeschript, C, C++, Latex
Databases: MySQL, PostgreSQL
Libraries: Numpy, Pandas, OpenCV, MLFlow,
Frameworks: PyQT, Django, Django-REST, React, Keras, TensorFlow, PyTorch, Huggingface, Apache Airflow, Minio
Tools & Technologies: Git, Docker, AWS, GCP, Lucidchart

Detail-oriented Software Developer and Researcher with professional experience in Backend Development, Server Side Programming, Machine Learning, Computer Vision, Image Processing.

Looking for opportunities to employ my knowledge and skills in research oriented tech-industries

Updates

2023

Started As A PhD Student At Indiana University Bloomington in Computer Science
Conference paper publishing of "GENERATING HUMAN ADMISSIBLE LANGUAGE FROM BENGALI SIGN LANGUAGE USING IMAGE PROCESSING" in 10th International Conference on Advanced Technologies

2022

Abstract paper "GENERATING HUMAN ADMISSIBLE LANGUAGE FROM BENGALI SIGN LANGUAGE USING IMAGE PROCESSING" accepted in 10th International Conference on Advanced Technologies
Conducted a seminar on Bangabondhu Olympiad 2022
Submitted Bengali Sign Language paper to IEEE Calcon
Orchestrated a workshop on ML & AI for Fourth Industrial Revolution

2021

Accepted paper on smart parking at IEEE Tensymp Region 10 Symposium 2021
Submitted Paper to IEEE Tensymp Region 10 Symposium 2021

Experience

Graaho

Software Engineer (ML)

Research and analyse regulatory documents and design database schema
Design and implement data crawling and data mining system
Develop Crawler Management Platform with Apache Airflow
Develope a computer vision based data parser platform
Implement NLP model(s) and system flow to automate structuring of unstructured data
Design and develop an AI Management Platform for renforce AI system
Platform: Python, Django, Pytorch, Scikit-Learn, Huggingface Transformers, NLTK, OpenCV, Apache Airflow

June 2021 - July 2023 | Reston, VA, USA

Graaho

Junior Software Engineer (ML)

Reseach and develop machine learning pipeline
Develop SOTA AI solutions for financial risk analysis
Research restaurant menu structure and develop automatic image based text extractor
Develop recommendation engine for komidaz, a food delivery solution of Graaho
Platform Python, Django, AutoML, Scikit-Learn, MindsDB, Tensorflow, Keras

Oct 2020 - May 2021 | Reston, VA, USA

Reseach Works

                
Smart Parking System
                  A smart parking management solution using computer vision
                
AccomplishmentsTools: Tensorflow, Keras, SSD mobileNet v2
Detect parking availability of a parking
Detect vehicle and its type and other important property
Store and send data through a CC camera to remote server

Bengali Sign Language Detection
                  Bengali sign language detection system using image processing
                
AccomplishmentsTools: Python Kivy, Tensorflow, Keras, OpenCV, Numpy
Prepared a dataset of Bengali hand sign
Detect and classify Bengali hand sign containing words
Provide both audio and visual feedback of detected hand sign

Covid-19 Time Series Analysis
                  A forecast for Covid-19 virus on live data
                
                Accomplishments
                Tools:Python, Scikit-Learn, Pandas, Numpy
Forecasted Covid-19 virus effects
Generated virus effect report
Predicted probable rate of spread of the virus

Face Expression Recognition
                  Detecting face expression for both human and animated figure
                
AccomplishmentsTools:Tensorflow, Keras, OpenCV, Numpy
Prepared dataset of face expression for both human and animated figures 
Created a classifier using VGG16 to detect face expression

Early Stage Heart Disease Detection
                  Detecting face expression for both human and animated figure
                
                Accomplishments
                Tools:Tensorflow, Keras, OpenCV, Numpy
Prepared dataset of face expression for both human and animated figures 
Created a classifier using VGG16 to detect face expression

Projects

                
Text Cluster Identification & Classification
                  A Computer Vision based text cluster detector and extractor
                
                Accomplishments
                Tools: OpenCV, AWS S3, PyMuPDF, Pytesseract, OCR
Identify text cluster from unstructured document
Classify clusters into segments
Extract detected cluster into raw text, passed to system for manipulation

Metadata Extraction From PDF Document
                  Extract metadata of pdf document from raw regulatory document
                
                Accomplishments
                Tools: Pytorch, Huggingface Transformer, PyMuPDF, MLFlow, Apache Airflow
Create a dataset of metadata collected from regulatory documents
Research and develop a metadata entity identifier and extractor
Developed a ML pipeline for model training and testing

Rule Recognition
                  Rule is a regulatory obligation which we have identified and extracted from regulatory documents
                
                Accomplishments
                Tools: Pytorch, Huggingface Transformer, PyMuPDF, MLFlow, Apache Airflow
Create a dataset of obligation and rules from regulatory documents
Researched and developed a POS based text classifier for rule recognition
Integrated to ML pipeline for easy train and deploy

Komidaz Recommendation Engine
                  A recommendation system for the Komidaz food delivery system
                
                Accomplishments
                Tools: Scikit-Learn, Numpy, Pandas, NLTK, GCN
Developed a recommendation service for the Komidaz application
Created a pipeline for restaurant and food based recommendation
Created continuous model training functionality for developing model based on data availability
                  
Qesfera
                  A credit risk modelling solution
                
                Accomplishments
                Tools: MindsDB, Scikit-Learn, Django-REST, Angular, Draft.js, Pandas, Seaborn
Generate dataset report and analysis visualization
Predict credit default probability
Automate ML training of classification model

JATSS Clawbot
                  A gesture controlled object collector
                
AccomplishmentsTools: MindsDB, Scikit-Learn, Django-REST, Angular, Draft.js, Pandas, Seaborn
Developed a hand gesture based controlled bot
Used Arduino to operate the robots and claws movement

Publications

10th International Conference on Advanced Technologies, NH Tonmoy, M Reza, N Ahmed, M Uddin, M Afridi, S Tanaz, N Huda Generating Human Admissible Language From Bengali Sign Language Using Image Processing
2021 IEEE Region 10 Symposium (TENSYMP) MI Iqbal, MI Leon, NH Tonmoy, J Islam, A Ghosh Deep Learning based Smart Parking For A Metropolitan Area