Hi, my name is

Mohammed Lokhandwala

An AI and Web Dev enthusiasist.

Innovating the world with AI and web development. I am a passionate developer who loves to create innovative solutions that make a difference. I am always eager to learn and explore new technologies.

Get In Touch

About Me

Hi there! I'm Mohammed Lokhandwala — a full-stack developer and AI enthusiast with a strong foundation in both web development and machine learning. I specialize in building modern, responsive applications using technologies like React, Node.js, and MongoDB. Alongside I specialize in applying AI to solve real-world problems — from computer vision and image enhancement to model tracking and automation. I'm always eager to learn, grow, and take on new challenges in tech.

Looking forward to sharpening my skills in a challenging environment. I guarantee enthusiasm in work and fulfilling my projects with inexpressible effort.

Experience

AI/ML Intern

SeePossible Innovative Solutions LLP • June 2024 - June 2024

  • Built and deployed an image upscaling API that lets users submit images with custom parameters (model type, resolution, size), secured with authentication —processed 5,000+ images with 99.9% uptime
  • Fine-tuned a deep learning super-resolution model specifically for jewelry imagery, achieving a 28% increase in image quality (measured via LPIPS and PSNR) improving clarity and sharpness while preserving details.
  • Used Weights & Biases for experiment tracking, model versioning, and performance visualization, enabling reproducible training workflows for both upscaling and interpolation pipelines.
  • Implemented a real-time quota and user usage tracking system using supabase, that supported 100+ concurrent users without latency spikes.
  • Developed a CLI based frame interpolation system to generate intermediate video frames from existing renders, eliminating the need for costly re-rendering—reducing post-production time and computational cost by nearly 50%.
  • Enhanced the interpolation model through architectural tweaks, hyperparameter tuning, and post-processing strategies and boosting frame coherence by 32% (based on temporal SSIM metrics).

Skills

Front-end

AngularAngular
HTMLHTML
CSSCSS
JavaScriptJavaScript
TypeScriptTypeScript
TailwindTailwind
ReactReact
streamlitstreamlit
ShadcnShadcn

Back-end

PythonPython
Node.jsNode.js
Express.jsExpress.js
FlaskFlask
fastAPIfastAPI
JqueryJquery
vitevite

Database

MongoDBMongoDB
PostgreSQLPostgreSQL
MySQLMySQL
SQLiteSQLite
ChromaDBChromaDB

Ops

GitGit
VercelVercel
NetlifyNetlify
Weight & BiasWeight & Bias

AI Framework & Languages

c++c++
TensorflowTensorflow
PytorchPytorch
KerasKeras
OpenCVOpenCV
Scikit-learnScikit-learn
PandasPandas
NumPyNumPy

Things I've worked on

Local Codebase Search Engine project image

Featured Project

Local Codebase Search Engine

  • Created a local code search engine to navigate large codebases efficiently without relying on external tools or services.
  • Wrote a custom OS-level file crawler with support for regex and semantic chunking to break code into meaningful segments.
  • Used MiniLM embeddings with ChromaDB to implement fast, vector-based code similarity search.
  • Added language-aware static analysis using tree parsing to improve search accuracy with symbol recognition and context-aware results.
ReactFlaskChromDBPython
Local ML Experiment Tracker project image

Featured Project

Local ML Experiment Tracker

  • Built a full-stack, offline ML experiment tracker like ML FLow to log, visualize, and compare training metrics, configurations, and results.
  • Developed a RESTful Node.js API for handling versioned uploads, metadata storage, and log processing, with a React dashboard using Recharts for real-time visualization.
  • Created a custom Python SDK for logging directly from script and CLI automatically without manual adding.
  • Implemented backend logic for config comparisons, metric aggregation, and automatic plot handling (Matplotlib/CSV/JSON), storing data in MongoDB and on local disk.
Node.jsPythonMongoDBRechartsReacttailwind CSStypescriptviteexpress
Billing and Inventory Management System project image

Featured Project

Billing and Inventory Management System

  • Developed a complete billing and inventory system with features like product cataloging, discount rules, stock management, and invoice generation.
  • Deployed at a local retail store and used daily for handling sales, inventory updates, and reports.
  • Built real-time inventory sync and role-based access (admin/cashier), with modular components for managing products,customers, and sales data.
  • Integrated thermal receipt printing and barcode generation from the UI to work smoothly with standard retail hardware.
Node.jsExpressMongoDBReactTailwind CSSjavascriptshadcn-uivite
End to End OCR Pipeline project image

Featured Project

End to End OCR Pipeline

  • Integrates CRAFT for precise text detection and CRNN for accurate text recognition, enabling comprehensive scene text extraction.
  • Achieves inference times as low as 60ms per image using PyTorch, making it suitable for applications requiring quick processing .
  • Employs CRAFT's character-level detection to effectively identify text in complex layouts, including curved and multi-oriented text .
  • Outputs results in structured JSON format, detailing bounding box coordinates and recognized text, facilitating seamless integration into various applications.
PyhtonPytorchOpenCVtorchvisionpillow
View More Projects

What's Next?

Get In Touch

Let's Chat! Whether you have a question, a project idea, or just want to connect, I'm always happy to hear from you. Drop me a message, and I'll be in touch soon!