SHAYM MOHAN
Manager - AI/ML, FeatherSoft Info solution Pvt
Hi, I’m currently working as Manager – AI/ML at FeatherSoft Info Solution Pvt. Ltd. With a strong passion for artificial intelligence and machine learning, I’ve spent 7 years exploring how intelligent systems can solve real-world problems—especially those that challenge logic, creativity, and problem-solving skills.
This website, PlaySolve, is a personal initiative to blend my professional expertise with my love for puzzles. It’s where code meets curiosity, and where ideas are transformed into working solutions—from Sudoku solvers and Rubik’s Cube algorithms to interactive block puzzle tools.
When I’m not solving puzzles with code, I’m usually exploring new trends in AI, mentoring teams, or building applications that bring innovative ideas to life.
CAREER OBJECTIVE
To leverage my expertise in Artificial Intelligence and Machine Learning to design innovative, scalable, and impactful solutions that solve real-world problems. I aim to lead and collaborate on cutting-edge projects that push the boundaries of intelligent systems—driving value for organizations and enriching user experiences, while continuously growing as a technology leader and mentor.
Thinking helps us move forward. Without thinking, people, organizations, and countries stop growing.
PROJECTS
Rubik's cube solving machine - Robotics
November 2012 – May 2013
The Rubik’s Cube Solving Machine is a robotics project that uses a mechanical arm and smart algorithms to solve a scrambled cube automatically. It showcases the integration of hardware control with intelligent logic. A perfect blend of engineering and problem-solving in action.
8x8x8 LED cube
February 2015 – May 2015
The 8x8x8 LED Cube is a 3D visual display made up of 512 LEDs, controlled to create stunning light animations. It demonstrates advanced electronics, timing control, and programming precision. A captivating fusion of creativity, code, and circuit design.
SUDOKU solver with c graphics
March 2009 – November 2010
The Sudoku Solver with C Graphics is a visually interactive program that solves Sudoku puzzles using custom-built logic and interface. This was the first algorithm I fully designed and implemented myself—marking the start of my journey into problem-solving with code. It combines classic puzzle-solving with low-level graphics programming in C.
Predict Watermelon Dimension
February 2023 – November 2023
This AI-based model is designed to predict the dimensions of watermelons using image data and machine learning techniques. It supports precision farming by helping farmers estimate yield and quality before harvest. A practical application of AI in agriculture, combining computer vision with real-world farming needs.
Predict SweetCorn Dimension
January 2023 – November 2023
This project delivers a custom AgriTech software solution to predict the size and dimensions of sweetcorn using image processing and AI. It enables farmers and suppliers to assess crop quality quickly and accurately without manual measurement. An innovative step toward smart agriculture, combining technology with real-world farming efficiency.
Puzzle Solving website(www.playsolve.in)
September 2011 – Present
www.playsolve.in is a web-based platform designed to solve and simulate logic-based puzzles like Sudoku, Rubik’s Cube, and block puzzles. It combines AI algorithms with interactive UI to provide both automated solutions and hands-on puzzle-solving experiences. The site also showcases my personal projects, reflecting my passion for coding, logic, and innovation.
4x4x4 LED cube
May 2013 – January 2014
The 4x4x4 LED Cube is a compact 3D light display built with 64 LEDs to create dynamic visual patterns and effects. It demonstrates the fundamentals of multiplexing, timing, and embedded control. A great blend of electronics and creativity, serving as a stepping stone to larger LED cube projects.
Smart Gate Control System - POC
February 2022 – April 2022
The Smart Gate Control System is a proof of concept (POC) that automates gate access using sensors and microcontrollers. It enhances security and convenience by enabling contactless entry based on predefined conditions. A practical IoT-based solution demonstrating efficient control and automation in smart infrastructure.
AWS DeepRacer — AI Self-Driving Car Experiment
November 2023 – February 2024
I conducted this experiment at Feathersoft to explore how reinforcement learning can be applied to autonomous driving. Using AWS DeepRacer, I trained a virtual self-driving car to navigate a race track by optimizing speed and control through trial and error. This hands-on project helped me understand real-world reinforcement learning concepts such as reward functions, policy optimization, and continuous model improvement — all within a fun and competitive environment provided by AWS DeepRacer.
Tissue Cell Phenotype Classification
January 2025 – April 2025
In this project, I developed an AI-based system to automatically detect and classify tissue cells based on their phenotypes. The model first predicts precise cell boundaries from microscopic images, and using those boundaries, a custom algorithm classifies each cell according to its morphological and structural characteristics. The key strength of this approach lies in its speed and scalability — the algorithm can accurately classify over 100,000 cells within just 15 minutes, making it highly efficient for large-scale biomedical image analysis and research applications.
Synthetic DNA Sequence Generator
May 2025 – july 2025
I built a robust, custom algorithm that generates synthetic DNA sequences on demand, allowing researchers to rapidly produce large batches of sequences by tuning a set of configurable parameters. The tool is optimized for speed and scale — users can adjust mutation rates, ATGC content, motif constraints, length distributions and other controls, then generate bulk datasets in seconds for simulation, benchmarking, or experimental design. Its flexibility and performance make it a powerful asset for labs and computational biology projects that need customizable synthetic sequence libraries quickly and reliably.
LED Array microscopy
January 2015 – December 2015
LED Array Microscopy is a biophotonics project aimed at digitally refocusing samples and reconstructing high-resolution images from low-resolution captures. It uses an LED array to illuminate the sample from multiple angles, with image shifts calculated for digital refocusing. The system combines electronics, Embedded C, and Arduino Uno to achieve precise control and image acquisition.
Cucumber Length Detector
November 2023 – June 2024
This project uses AI-based image processing to automatically detect and measure the length of cucumbers with high accuracy. Developed as an Android application, it allows users to capture or upload an image and manually outline the object to determine its length. While the primary focus was on cucumbers for agricultural use, the system can be adapted to measure any object’s length, making it a versatile tool for real-world measurement tasks powered by AI.
AI Chatbot Framework
January 2024 – August 2024
This project focuses on building a powerful and flexible chatbot framework using LangChain and LLMs such as ChatGPT, Gemini, and local models. I developed various types of chatbots — including RAG-based chatbots for documents, SQL databases, graph databases, and Excel data — as well as domain-specific assistants for doctors and company revenue analysis. The framework is designed for easy deployment on any server and allows users to configure their chatbot’s template, personality, and behavior through a simple interface. Once configured, the chatbot can be seamlessly integrated into web or mobile applications using a lightweight snippet, making it an adaptable solution for diverse use cases.
Image Background Remover
January 2023 – June 2023
This project uses advanced computer vision and deep learning techniques to automatically remove the background from images with high precision. The system intelligently detects the main subject and separates it from the background, producing clean and transparent outputs suitable for design, e-commerce, and photo editing applications. Built for speed and accuracy, this tool can process multiple images in seconds and supports fine-tuning for edge refinement, shadow preservation, and custom background replacement.
Leaf Disease Detector
February 2019 – September 2019
This Android application uses AI-powered image analysis to automatically detect and identify diseases in plant leaves. By capturing an image through the mobile app, the system processes it using a trained deep learning model that recognizes patterns, textures, and color variations associated with common plant diseases. The app then provides instant feedback on the detected disease, helping farmers and researchers make quick, informed decisions. Designed for accessibility and accuracy, this tool brings the power of AI-based plant health monitoring directly to smartphones.
© PLAYSOLVE.in All right reserved.