Experience

• 

  Techno teams

  Apr 2016 - Feb 2017

  Frontend web developer

  - Proficiently designed web pages using HTML, CSS, and JavaScript, ensuring visually appealing and functional user interfaces.- Demonstrated expertise in customizing WordPress websites, tailoring them to meet specific client requirements and design preferences.- Skillfully modified PHP scripts for web scraping purposes, enhancing data extraction efficiency and accuracy.- Developed and implemented web scrapers using the Python stack, enabling automated data collection and analysis.- Contributed to web development and data extraction projects, showcasing a versatile skill set in various aspects of web and data-related tasks. Show less

• 

  American international university-bangladesh

  Sept 2017 - Dec 2018

  Research assistant

  AIUB Data Science Lab Categorized Affect Map - Conducted extensive research and development, resulting in the creation of the GIS application, "Categorized Affect Map" which offered valuable insights into spatial data and emotions. - Explored various GIS research and crowd-sourcing concepts, implemented auto-feature engineering techniques for machine learning models, and took on the role of thesis group manager to ensure the project's successful execution.Bengali WordNet - Led and supervised a dedicated thesis group in the experimentation and testing of features within the Bengali WordNet, contributing to the enhancement of this linguistic resource. - Demonstrated effective project management skills while overseeing the research and experimentation processes, ensuring the successful completion of the project's objectives.Lab link: https://www.researchgate.net/lab/AIUB-Data-Science-Lab-Tabin-Hasan Show less

• 

  Graaho technologies

  Feb 2019 - now

  [I worked as the Renforce Work Description from October 2021 to January 2023.]- Team leadership: Leading backend & ML team for client and in-house projects.- App design & architecture: Architecting Django & Django Rest Framework-based apps. - Data-informed decisions: Making data-driven design and development decisions. - Backend development: Developing robust backend solutions. - ML integration: Leveraging ML for enhanced project outcomes. - Effective communication: Communicate with AWS Saas Factory and CaraHost stakeholders.AlgoRec: Researched and developed two recommendation models utilizing the Nvidia Merlin framework. Additionally, led the backend, machine learning, MLOps, and data engineering efforts to create a versatile recommender system platform delivered as a SaaS application.• Platforms: Pytorch, Scikit-learn, MLflow, Apache Airflow, Django, DRF, Celery, Redis, Postgres, Docker.FaceID Auth: Designed and developed a facial recognition-based authentication microservice using the PyTorch framework for the GCognito authentication module, enabling seamless authentication and authorization.• Platforms: Pytorch, Scikit-learn, OpenCV, FastAPI, Celery, Redis, Postgres, Docker.Traffic AI: Developing a traffic analysis and toll collection system, leveraging the YOLOv8 computer vision model, annotation library supervision, and real-time data processing through Apache Kafka in the pipeline.• Platforms: Pytorch, Supervision, YOLOV8, Apache Kafka, WebRTC, Celery, Redis, Postgres, Streamlit.Gaze AI: Developed a gaze estimation system using MediaPipe, OpenCV, annotation library supervision, and real-time data processing with Apache Kafka. The system tracks participant’s eye movements in response to product displays. It is in alpha release for a supermarket chain to measure attention and engagement levels for market research.• Platforms: MediaPipe, Supervision, OpenCV, Apache Kafka, WebRTC, Celery, Redis, Postgres, Streamlit. Show less - Led backend and ML engineering teams for client and in-house projects, ensuring successful execution.- Architected, designed, and developed Django & Django Rest Framework apps for enhanced functionality.- Provided valuable data-driven insights while making critical design and development decisions.- Spearheaded data-centric strategies to optimize project outcomes and client satisfaction.- Focused on effective design and tech solutions to drive project success.Qesfera: Designed and developed a financial machine learning application for credit risk modeling, incorporating an interactive dataset visualizer, prediction analyzer, and reporting features within the platform.• Platforms: Mindsdb, Pytorch, Scikit-learn, MLflow, Django, DRF, Celery, Redis, Postgres, Docker.KhaoDao Recommendation Engine: Incorporated real-time prediction capabilities into our DLRM-based implementation, focusing on achieving the lowest possible latency.• Platforms: Tensorflow, Scikit-learn, Django, DRF, Celery.KhaoDao Menu Digitization AI: Led the development of a Windows-based annotation tool that uses a PyTorch-based OCR implementation for both Bengali and English languages.• Platforms: Pytorch, Opencv, EasyOCR, Windows, PyQT.KhaoDao KYC: Developed a computer vision application called KhaoDao KYC, a PyTorch-based OCR implementation supporting both Bengali and English languages, and provided it as a web-based service.• Platforms: Pytorch, Opencv, EasyOCR, Linux, Django, DRF, Celery, Docker.BengalMeat: Developed a livestock commerce platform for Bengal Meat to streamline procurement, inventory management, and sales processes for their Qurbani cattle, all within a single platform.• Platforms: Django, DRF, Celery, Postgres, Nginx, Docker, Ansible. Show less - Conducted R&D for computer vision and ML models.- Created analytics reports for KhaoDao.- Delivered presentations on in-house applications to clients and events.Smart Guess: Developed a computer vision application using the instance segmentation technique (Mask R-CNN) to identify cow girth coordinates, both front and back. Utilized Keras functional API’s multi-input-output network architecture to predict height, length, weight, and breed.• Platforms: TensorFlow, Keras, Scikit-learn, Linux, DRF, Celery, Redis.• Achievement: Basis National ICT Award 2019 for AI in Agrotech.Dell Microsoft License Management Portal: Developed Microsoft License Management Portal for Dell, streamlining software purchases, license management, and budget planning. Implemented features for online ordering, access control, and reporting, improving user experience.• Platforms: Django, Celery, Postgres, Minio, Apache.Qesfera Credit Risk Modeling: Developed a prototype for a financial machine-learning application focused on credit risk modeling. Experimented with automated feature engineering techniques and implemented best practices.• Platforms: Scikit-learn, Pandas, Tensorflow, Keras, MindsDB, Pycaret.KhaoDao Recommendation Engine: Researched and prototyped a contextual recommendation model using DLRM.• Platforms: TensorFlow, Keras, Scikit-learn, Linux.KhaoDao BI Report: Designed analytic reports in Power BI using customer data (both quantitative and qualitative, market and internal) for the business team.• Platforms: PowerBI. Show less

  • Senior Software Engineer (ML/DE/Backend/Cloud)

    Oct 2021 - now

  • Software Engineer (DS/ML/Backend/Cloud)

    Mar 2020 - Sept 2021

  • Junior Software Engineer (DS/ML/Backend/Cloud)

    Mar 2019 - Feb 2020

  • Software Engineer

    Feb 2019 - Feb 2019
• 

  Renforce

  Oct 2021 - Jan 2023

  Senior software engineer (ml/de/backend/cloud)

  - Built ML & Data Regulatory Solution: Developed a cutting-edge regulatory solution during the Alpha Phase.- ML Model Development: Created advanced ML models and established efficient MLOps workflows.- Architectural Design: Designed distributed system middleware, workflow manager, ML lifecycle processes, Django-based applications, and more.- Effective Communication: Maintained clear communication with project and product managers to align objectives.- Led and managed a multidisciplinary team of Software, ML, and Automation engineers for project success.Text Cluster Classifier: Developed a machine learning model to identify and categorize text clusters from PDF documents. The model structures the text as obligations, which are then verified using ontology modeling before being ingested into the system.• Platforms: Pytorch, Scikit-learn, MLflow, FastAPI, Celery.Renforce Workflow Manager: Designed a workflow manager for multi-step workflows that consume events from Apache Kafka, creating dynamic DAGs from templates or executing other tasks on DAGs using Apache Airflow.• Platforms: Apache Airflow, Apache Kafka, Kafka-python, Docker.Renforce Robotics Management Platform: Designed a platform for creating robots and automating robotic tasks that initiate workflows using a workflow manager. This platform handles dynamic instructions and parameters to crawl, scrape, extract, and ingest data into the data lake.• Platforms: Apache Airflow, Apache Kafka, Selenium, Minio, Django, DRF, Celery, Redis, Postgres, Docker.Renforce AI Management Platform: Designed a platform for managing and experimenting with renforce models, including model training, development, deployment, and monitoring.• Platforms: Apache Airflow, Apache Kafka, MLflow, FastAPI, Minio, Django, DRF, Celery, Redis, Postgres, Docker.Product Website: https://www.renforce.io/Company Profile: https://www.linkedin.com/company/renforce-io/Updated Research Works - https://www.renforce.io/research Show less