DAV Model School

Indian Institute of Technology, Kharagpur

Udacity

D.A.V. Public School, Midnapore

Machine Learning Engineer, Ideator, Futurist, Cosmophile  -  IIT KGP

Hyderabad, Telangana, India

Jukshio

Staff Scientist

Created Microservices based scalable data pipeline deployed to AWS cloud for ML product, using pub-sub messaging.• Deployed Dockerized data processing system: ingestion, cleaning, transform; for live incoming data using AWS Cloudformation• Troubleshooted and achieved 99.9% data integrity• Achieved max throughput of 10000 ETL tasks per 5 minutes

Fission Labs

Software Engineer

Solved real-world business problems utilising statistical analyses and visualizations, developing models using feature engineering and machine learning techniques like SARIMAX, Multiple Linear Regression, Random Forest, XGBoost, Clustering, et al., optimizing data pipeline, deploying end-to-end solution into production on Dataiku platform.• Improved efficiency by 10% in loading-unloading of FMCG stock, using extensive Feature Engineering and multi-step Ensemble Modelling.• Achieved 70% accuracy in projections of promotions' effectiveness on B2B sales for biscuit brands; collaborated with project members to analyze patterns and evaluate candidate Machine Learning models.• Improved manageability, reduced runtime by factor of 10, of client's data pipeline for US stocks' Short Interest forecasts, using modular & optimal Python numpy, pandas , sklearn based ML flow.• Streamlined reusability across projects by building a modular & optimized codebase for sliding-window Time series Forecasting, independent of Machine Learning algorithm used.

Alumnus Software Limited

Engineer, Data Science Team

- Implemented proof-of-concept stress incident detection Chatbot project consisting Android app, Python Django Server and Google Dialogflow NLP agent.• Trained a comprehensive Dialogflow bot with custom user intents leveraging application specific data.

Trainee Engineer

Research Assignment at CNERG group at Department of Computer Science and Engineering.

Research Assistant

Worked in the <a href="http://Bing.com">Bing.com</a> Team:• Automated Feature Engineering for existing Debugging tool for location query results• Built Machine Learning module on Big Data platform to determine most important locations for an application, from past search logs

Microsoft India

Machine Learning Intern

Worked for National Digital Library, IIT Kharagpur, by MHRD, India as part of B.Tech Project:• Engineered prototype for web portal, with active feedback from mentor, for searching and accessing previous years' Question Papers of Board Exams and Entrance Examinations, single-handedly.  - Used Django Frameworks and MariaDB database backend.• Automated metadata creation for indexing new educative content files in the Digital Library usingPython Information Retrieval program making use of Open-source Python modules such as lxml, ebooklib.• Achieved automation of conversion of a large batch of metadata from one format to another.

National Digital Library

Project Student

Interned at Testing Automation Department in collaboration with batchmate, with Proprietary software.• Integrated SAP tool with third-party software suite using C# prototype Windows application.• Experimented on prototype Android application for testing mobile apps as a Proof of Concept.

Intern

Designed and developed visual simulations of mechanics topics including Simple Harmonic Motion using Processing libraries and IDE, proportional to real-time motion characteristics for e-learning purposes.

thelearningpoint.net

Freelancer

Flash

Flash Developer

Collaborated with senior to build an online Business Simulation Game site using HTML, CSS and PHP.

Intaglio - IIM Calcutta International B-school Meet

Project Intern

Team AGV is multidisciplinary research & development group in the field of Robotics & Artificial Intelligence aimed at building fully operational self-driving car.• Acquainted with various techniques A* search, Simultaneous Location and Mapping (SLAM),Reinforcement learning and Imitation Learning on 3D point cloud from LIDAR, stereoscopic vision.• Contributed to SONAR and GPS sensor modules for path-planning and obstacle avoidance using Kalman Filter in Version 1 of bot, Eklavya, team's entry to IGVC '12, Michigan, USA.• Collaborated on simulating motion dynamics using Gazebo Robotics simulator and Robot Operating System (ROS) for Version 2 of bot participating in IGVC ‘13 and for Lunar Autonomous Rover prototype chassis for Team Indus, Google Lunar Xprize finalist.

Autonomous Ground Vehicle Research Group

Developer

Worked on and addressed the problem statement within a couple of weeks :• Basic Shape Detection from camera video input by a differential drive robot, using OpenCV libraries in VC++ Visual Studio Project

Technology Robotix Society

Image Processing Winter Workshop Participant

Some coding projects: "​ <a href="http://github.com/enlighter">github.com/enlighter</a> "​Project: Semantic Digital LibraryAcademic Projects:• Used Hadoop Streaming utility to extract and analyze meaningful information from large website and sales logs, with Python mapper and reducer scripts to run MapReduce on a Hadoop cluster.• Built a prototype of a Online Course System website using Django and MariaDB• Used C++ 11 for Computer Graphics Projects. Worked as part of Autonomous Ground Vehicle Research GroupBeen a member at Robotics group KRAIG, RobotixWon 'Best freshers'​ award'​ in Robotics competition implementing Image Processing called 'Robocop'​ in Kshitij 2011Extra-curricular Activities:- Awarded one of the best volunteers for the year at NSS in capacity of Deputy Group Leader.- Participated in IIT KGP’s various Technical and Social & Cultural Inter Hall Activities.- Hobbies include Sketching, road trips, trekking and Crypto-currencies.

Sushovan Mandal