Skip to content
View imadhajaz's full-sized avatar

Block or report imadhajaz

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
imadhajaz/README.md


πŸ‘‹ Hey there, I'm Imadh Ajaz Banday

πŸ§‘β€πŸ’» About Me

  • πŸŽ“ Graduate Student, University of Texas at Arlington (UTA), MS in Computer Science (GPA: 4.0β€”Aug 2024 to May 2026)

  • πŸŽ“ Bachelor of Engineering (Computer Science), B. M. S. College of Engineering, Bangalore, India (GPA: 3.5 β€” Aug 2020 to June 2024)

  • πŸ’» Machine Learning Engineer Intern, Vensora Inc., Dallas, TX, United States (Remote β€” Sep 2025 to Dec 2025)

    β†’ Scaled the platform into a production-ready AI system for automated code generation, reducing manual engineering effort by ~70%

    β†’ Built multi-step, asynchronous AI pipelines, improving system reliability, observability, and workflow orchestration

    β†’ Developed AI agent-based validation and self-correction loops, increasing successful generation rates by ~3x

    β†’ Designed agent memory mechanisms leveraging relational (PostgreSQL) and object storage (S3) to track progress, maintain context, and enable stateful multi-step workflows

    β†’ Implemented MLOps-driven improvements, including monitoring, validation layers, and iterative feedback pipelines for consistent AI outputs

    β†’ Improved system robustness through structured outputs and schema validation, significantly reducing failure cases

    β†’ Optimized end-to-end generation performance and consistency for real-world usage and large-scale workflows

    β†’ Completed system deployment aligning it for enterprise production use

  • πŸ’» Data Science & AI/ML Engineer Intern, Vensora Inc., Dallas, TX, United States (Remote β€” June 2025 to Aug 2025)

    β†’ Led development of an AI-powered platform for automated application and code generation from structured requirements

    β†’ Built LLM-driven code generation workflows, transforming high-level inputs into full-stack application components

    β†’ Developed context-aware (RAG-based) AI systems, improving output accuracy and reducing hallucination-related failures by ~40%

    β†’ Designed core system architecture and service interfaces, enabling scalable interaction between AI services and backend systems

    β†’ Implemented persistent context and memory layers using PostgreSQL and object storage (S3) to maintain structured state and support context-aware generation across workflows

    β†’ Established early MLOps practices for LLM pipelines, including reproducible workflows, evaluation loops, and structured output handling

    β†’ Collaborated across design and engineering to align AI-generated outputs with real product workflows and requirements

    β†’ Actively contributed in Scrum meetings and Agile sprints, collaborating cross-functionally to deliver GenAI features

    β†’ Led foundational engineer role in defining the technical direction and foundational architecture of an early-stage AI product

  • πŸ’» Software Engineer Intern, Velozity Global Solutions

    β†’ Worked on the deployment of a full-stack solution by integrating front-end and backend services with Express.js, architecting a scalable microservices architecture that improved application load times by 15% for key user-facing pages.

    β†’ Implemented automated API testing protocols using Postman and orchestrated CI/CD pipelines to ensure continuous integration, reducing manual testing effort and improving API response times by an average of 30 milliseconds across all endpoints.

  • πŸ”§ Tech Lead, CodeIO, BMSCE
    β†’ Mentored 10+ junior developers and expanded modules in the BMSCE student-faculty ERP portal, improving usability and performance.

  • πŸ€– Conducted research using Hugging Face Transformers, Streamlit dashboards, and vector stores like ChromaDB, FAISS, and Pinecone for vision-language modeling and retrieval pipelines.


πŸš€ Featured Projects

πŸ’Έ MoneyMind

React Native Β· Node.js Β· Firebase Β· Plaid API
β†’ An intelligent personal finance tracker categorizing transactions with 90%+ accuracy and providing smart budget insights.

β˜• CoffeeCare

Python Β· Flask Β· YOLOv8 Β· OpenAI Β· LangChain Β· RAG Β· ChromaDB
β†’ Real-time coffee leaf disease detection app with OCR and an LLM-based retrieval QA pipeline for farmer support. Integrated vector search via ChromaDB.

Python Β· OpenCV Β· CNN Β· Transformers
β†’ Built a hybrid CNN-transformer pipeline to detect tampered videos using subtle facial shifts and temporal inconsistencies.

Flutter Β· Dart Β· Firebase Β· Firestore
β†’ A secure real-time healthcare chat app featuring typing indicators, presence tracking, and sleek material UI.

E-commerce Β· React Β· Node Β· Delivery Intelligence
β†’ Implemented a smart multi-delivery cart split system by zip code, with unified payment processing across sub-orders.


πŸ“š Publications

  1. πŸ“„ Overcoming LLM Challenges using RAG-Driven Precision in Coffee Leaf Disease Remediation
    IEEE ICETCS 2024 β€” DOI: 10.1109/ICETCS61022.2024.10543859

  2. 🧠 Vision Encoder-Decoder Models for AI Coaching
    IEEE INOCON 2024 β€” DOI: 10.1109/INOCON60754.2024.10512280

  3. 🌍 Language Detection for Transliterated Content
    IEEE 2024 β€” DOI: 10.1109/IEEECONF58110.2023.10520601


🧰 Tech Stack

πŸ’» Programming Languages

πŸ€– AI & Prompt Engineering

πŸ“Š Data & Analysis

🌐 Web & Backend

πŸ› οΈ Tools & Platforms

Popular repositories Loading

  1. DeepFake-Detection DeepFake-Detection Public

    DeepFake detection is a data science project aimed at detecting potential deepfakes in a video.

    Jupyter Notebook 2

  2. DeepFake-Detection-Website DeepFake-Detection-Website Public

    A full stack website developed to test videos to identify any potential deepfakes in the video.

    HTML 1

  3. imadh-a imadh-a Public

    Config files for my GitHub profile.

  4. Fashion-Fiesta Fashion-Fiesta Public

    E-Commerce Website

    HTML

  5. JAVA-LAB-BMSCE JAVA-LAB-BMSCE Public

    Java

  6. ADA-LAB-BMSCE ADA-LAB-BMSCE Public

    C