Loading Portfolio...
Bharat Kaurav

Bharat Kaurav

|

Software Engineer at Metzev • AI Researcher • Full Stack Developer •

15K+
Users Impacted
$50K+
Revenue Generated
5
Major Projects
1
Papers Published
View Resume Explore Projects

Current Focus

Software Engineering

Building scalable EV battery management systems at Metzev

AI Research

Published research in computer vision and flood prediction

Education

B.Tech CSE at IIT Indore (8.44 CGPA)

Tech Stack Highlights

AI/ML

PyTorch Transformers Computer Vision

Full Stack

Next.js React Python

Cloud & DevOps

AWS Docker CI/CD

Professional Experience

Software Engineer

Metzev May 2025 – Present Georgia, USA (Remote)
  • Developed a scalable EV battery management platform using Next.js and Supabase with real-time data handling
  • Implemented performance optimizations including code splitting, lazy loading, and database indexing
  • Scaled platform to 15,000+ users across 1,000+ stores in USA, securing 5 major clients
  • Contributed to $50K+ revenue growth through platform optimization and client acquisition initiatives
  • Ensured platform reliability through performance-focused design and robust backend architecture

Machine Learning Engineer

unstudio.ai Nov 2024 – Apr 2025 Gurgaon, India (Remote)
  • Engineered a scalable, personalized product training and visualization pipeline for FLUX using LoRA fine-tuning
  • Improved image quality and experience for 2000+ users globally, driving customer retention
  • Built distributed training system using Redis Pub/Sub with robust error handling and automated CI/CD via GitHub Actions
  • Optimized model inference reducing costs by 30%, contributing to improved profit margins

Research Intern

LIPG, IIT Indore Jul 2023 – Present Indore, India
  • Created a 5000+ image dataset using DINOv2 and SAM for automated segmentation mask generation
  • Achieved 96% SOTA classification accuracy using ECA, SAM, hybrid CNN-Transformer architecture with frequency feature interaction
  • Research paper published in Sustainable Cities and Society journal (Impact Factor: 10.5)

Machine Learning Intern

Styldod Inc. Jun 2024 – Oct 2024 Bangalore, India (Remote)
  • Optimized diffusion model inference using Onediffusion and Stablefast, achieving 50% faster inference and 30% GPU usage reduction
  • Developed scalable pipeline for image collection, quality control, and labeling; tested with 300k+ images

Key Collaborations

Virtual Try-On Optimization

Alle • Dec 2024

Led team of 5 to optimize LoRA-based try-on system for FLUX, achieving 2.5× faster boot-up and 98.7 PSNR on A40/H100 GPUs

LLM-SQL Generation

Attentions.ai • Apr 2024

Led team of 5–7 to fine-tune LLaMA, Gemma, Qwen & Phi models for complex SQL query tasks; achieved 90% accuracy

Featured Projects

Sobra: Soybean Disease Classification & Explanation

Apr 2025 • Latest

Multimodal AI system for soybean disease classification using vision-language models and attention-based reasoning, achieving 95% accuracy on real-world data. Integrated RAG pipeline for symptom-based remedy suggestions from damaged leaf images.

Vision-Language Models RAG PyTorch Computer Vision Explainable AI

Dynamic Agentic RAG System

Inter IIT Tech Meet 13.0

Pathway-integrated dynamic agentic RAG system with real-time data ingestion, tool calling, and modular tool integration. Outperformed closed and open-source LLMs in knowledge-intensive tasks.

Python RAG Knowledge Graphs Docker Streamlit

Behavior Simulation Challenge

Bronze Medal • Nov 2023

Tweet prediction model with 96% accuracy and fine-tuned LLaMA for engagement optimization tasks. Processed 1M+ multimodal records (text, image, audio, video).

LLaMA NLP Multimodal Processing Deep Learning

Automatic TikTok Generator

May 2024

Automated video editing pipeline achieving 50-60s processing for 5s UHD videos. Complete pipeline with AnimateDiff and ControlNet, evaluated using SOTA benchmarks.

AnimateDiff ControlNet Video Processing Computer Vision

Hostel & Mess Management System

IITI Academic Office

Complete Django-based mess and hostel allocation system for IIT Indore. Engineered full-stack solution with comprehensive documentation and security measures.

Django Python Database Design System Architecture

Publications

An explainable deep neural network with frequency-aware channel and spatial refinement for flood prediction in sustainable cities

Sustainable Cities and Society, Vol. 130, 2025
Authors: B. Kaurav, S. S. Dar, A. Jain, C. S. Raghaw, M. Z. U. Rehman, N. Kumar

Developed an explainable deep neural network architecture incorporating frequency-aware channel attention and spatial refinement mechanisms for accurate flood prediction in urban environments, contributing to sustainable city planning and disaster management.

View Publication

Technical Arsenal

AI/ML & Deep Learning

Core frameworks and libraries used across research and production systems

  • PyTorch: Production-scale model development, FLUX LoRA fine-tuning, CNN-Transformer architectures
  • Transformers: Vision-language models, LLaMA fine-tuning, multimodal processing pipelines
  • Diffusers: Stable Diffusion optimization, inference acceleration, custom pipeline development
  • Research Methods: Frequency-aware channel attention, spatial refinement, explainable AI
PyTorch Transformers Diffusers TensorFlow Scikit-learn Pandas & NumPy MLflow Weights & Biases LangChain Optuna

Computer Vision & Research

Published research in flood prediction, agricultural AI, and generative systems

  • Research Paper: Frequency-aware CNN-Transformer for flood prediction (Impact Factor: 10.5)
  • Vision Models: DINOv2, SAM, YOLO for segmentation and object detection
  • Generative AI: FLUX LoRA, ControlNet, AnimateDiff for production applications
  • Agricultural AI: Disease classification with 95% accuracy, explainable recommendations
OpenCV DINOv2 SAM FLUX LoRA Stable Diffusion ControlNet YOLO Face Recognition AnimateDiff PEFT & LoRA

Full Stack Development

Production systems serving 15K+ users with scalable architectures

  • EV Platform (Metzev): Next.js, Supabase, real-time data handling for 1000+ stores
  • Portfolio Platform: GraphQL, TypeScript, OAuth integration, cloud image processing
  • Academic Systems: Django-based mess/hostel management for IIT Indore
  • Performance: Code splitting, lazy loading, database optimization
Python JavaScript/TypeScript Next.js React Django FastAPI Supabase GraphQL SQL PostgreSQL Next Auth MongoDB

RAG & LLM Systems

Advanced retrieval systems and large language model applications

  • Dynamic RAG: Pathway integration, real-time data ingestion, tool calling
  • LLM Fine-tuning: LLaMA, Gemma, Qwen models for SQL generation (90% accuracy)
  • Knowledge Graphs: Modular tool integration, evaluation pipelines
  • Agricultural RAG: Symptom-based remedy suggestions, explainable recommendations
RAG Pipelines LLaMA Knowledge Graphs Pathway Vector Databases LangChain Embedding Models Chroma

DevOps & Infrastructure

Production deployment and scaling for high-traffic applications

  • Containerization: Docker deployment, CI/CD pipelines, automated testing
  • Distributed Systems: Redis Pub/Sub, error handling, monitoring
  • Cloud Platforms: AWS deployment, cost optimization (30% reduction)
  • Performance: Load balancing, caching strategies, database indexing
Docker GitHub CI/CD Redis AWS Linux Git Nginx Kubernetes Monitoring

Data Science & Analytics

Large-scale data processing and behavioral analysis

  • Multimodal Data: 1M+ records processing (text, image, audio, video)
  • Behavioral Analysis: Tweet prediction with 96% accuracy, engagement optimization
  • Dataset Creation: 5000+ image datasets, automated annotation pipelines
  • Evaluation: SOTA benchmarks, PSNR analysis, statistical validation
Pandas & NumPy Matplotlib Seaborn Data Preprocessing Statistical Analysis Plotly Feature Engineering Model Evaluation

Education

B.Tech. in Computer Science & Engineering

Indian Institute of Technology Indore 2022 – 2026 (Expected) CGPA: 8.44

Key Courses

Data Structures & Algorithms Database Systems Software Engineering Computer Networks Operating Systems System Architecture Parallel Computing Linear Algebra Machine Learning Computer Vision

Achievements

Bronze Medal

12th Inter IIT Tech Meet, Adobe's Behaviour Simulation Challenge (2023)

JEE Advanced

All India Rank 1083 (2022)

JEE Mains

All India Rank 5995 (2022)