AI Product Builder · LLM Application Developer

Ayan
Jawaid

AI Product Builder and Generative AI Developer turning ideas into working, shipped products. I combine technical knowledge with product thinking across LLM applications, RAG pipelines, and agentic workflows.

View Projects NinthLoop Studio Resume GitHub LinkedIn Blog
Ayan Jawaid

Achievements & Recognition

Guinness World Record Holder

Most users in an AI video lesson (14,075 participants). Kanz & Saudi Arabia's Ministry of Human Resources.

July 2026

Gold — Kanz AI Global Hackathon

Project: NexoRaid: Privacy-First AI Workspace. AI Showcase Featured Project. Certified by LAU & Saudi Arabia's Ministry of HRSD.

Kanz Portfolio

Published Research Paper

"Performance Analysis of Machine Learning Algorithms for Brain Stroke Prediction" presented at a National e-Conference.

View Paper

Personal Projects

A curated showcase of local-first AI assistants, mobile chat companions, and innovative AI tools built by me.

View GitHub
VoxLens Thumbnail
Next.js Web Speech API Groq & Gemini

AI-Powered Pronunciation Checker

VoxLens

Built an AI-powered English pronunciation checker supporting both uploaded audio files and in-browser microphone recording via Web Speech API.

  • • Designed a flexible multi-provider speech-to-text architecture — auto-routes across OpenAI, Groq, Gemini, and OpenRouter.
  • • Implemented configurable AI feedback pipeline using a separate coaching LLM, decoupled from transcription provider.
  • • Built a privacy-first system: no audio, transcripts, or scores are stored — results remain browser-session only.
Live Demo
NexoRaid Thumbnail
WebLLM WebGPU RAG

Privacy-First AI Workspace 🥉 Kanz AI Bronze

NexoRaid

Built an AI workspace running open-source LLMs (Llama 3.2, Qwen 2.5) 100% in-browser via WebGPU — zero cloud dependency or API key required.

  • • Smart Provider Router auto-switches between local WebLLM inference and cloud APIs based on network state and privacy mode.
  • • Integrated browser-native OCR (Tesseract.js), interactive PDF viewer + text extraction, and agentic web search with real-time RAG.
  • • Ephemeral session architecture mathematically wipes all credentials and local storage on timer expiry — privacy-first by design.
Live Demo
Cars24 Price Prediction App Thumbnail
Python Streamlit Scikit-Learn

Machine Learning Regression App

Cars24 Price Prediction App

A simple yet powerful Streamlit web app that predicts the resale price of used cars using a machine learning regression model trained on Cars24 dataset.

  • • Estimates resale value based on features like Kilometers Driven, Engine Power, Fuel, and Transmission.
  • • Displays the predicted price in Lakhs (₹) and implements a smart car suggestion logic to show the top 10 similar cars within ±1 Lakh of that value.
Live App

My Designs

Concept UI/UX designs and visual explorations.

Porsche Concept

Porsche Concept

Luxury Watch Concept

Luxury Watch Concept

Client Projects

Real-world client engagements involving OTT infrastructure, cloud architecture, backend systems, and enterprise platforms.

Client Engagement • Jan 2026 - June 2026

E-Commerce CMS Platform

Freelance AI & Backend Engineer

Contributing to the development of a highly scalable, enterprise-grade e-commerce content management system designed to streamline online retail workflows and boost operations.

  • • Optimizing administrative workflows for massive scale operations, handling complex **Product and Inventory Management** systems.
  • • Architecting high-concurrency **Order Processing pipelines** and resilient background workflows.
  • • Designing and documenting robust, secure **Backend APIs** for smooth data flow and client platform integrations.
Next.js ChatBot Node.js PostgreSQL Docker REST APIs
Client Engagement • Feb 2025 – Dec 2025

VIVLY OTT Platform

Freelance AI & Backend Engineer

STREAMFORGE is a custom OTT media processing and distribution platform built to automate video ingestion, transcoding, storage, and global delivery.

Key Features & Deliverables
  • • Adaptive bitrate HLS generation for seamless multi-device streaming.
  • • FFmpeg-based processing pipelines for high-efficiency transcoding.
  • • AWS S3 secure media storage & raw ingestion workflows.
  • • CloudFront CDN integration for low-latency global delivery.
  • • Dockerized deployment orchestrating scaling and containerization.
  • • Detailed analytics and robust job tracking dashboard.
AWS S3 CloudFront FFmpeg HLS Streaming Docker Analytics

Experience

iMerit

2025 - PRESENT

AI Training Associate (Freelance)

  • • Evaluated AI-generated outputs using structured rubrics and quality guidelines to improve LLM model performance across diverse task types.
  • • Performed preference ranking, comparison tasks, and reasoning-based assessments across diverse LLM outputs.
  • • Worked with multimodal datasets (image, text, contextual) and participated in calibration sessions for model training pipelines.

Freelance AI & Product Engineer

FEB 2025 - PRESENT

Client Projects

E-Commerce CMS Platform
Jan 2026 - June 2026
  • Building a scalable CMS covering product management, inventory, order processing, and admin workflows with a focus on API integration and performance.
  • Implementing role-based access control, real-time stock management, and multi-payment gateway integration using Django REST Framework and React.
STREAMFORGE (VIVLY)
Feb 2025 – Dec 2025
  • Designed and built an enterprise-grade OTT video processing and distribution platform powering media workflows for the VIVLY ecosystem.
  • Built automated video ingestion, transcoding, adaptive bitrate HLS generation using FFmpeg (360p–4K), and cloud distribution pipelines.
  • Integrated AWS S3 + CloudFront for scalable global storage and delivery; containerised with Docker on AWS infrastructure.

Feynn Labs

JUL 2024 - SEP 2024

Data Science Intern

  • • Built Tableau and Excel analytical dashboards for client reporting; implemented RFM modelling on 50,000+ record datasets to identify high value customer segments.
  • • Validated and cleaned large structured datasets; produced structured KPI reports that improved stakeholder decision-making turnaround by 10%.

Techplement

JUN 2024 - JUL 2024

Python Developer Intern

  • • Automated Excel-based reporting workflows using Python, reducing manual effort by 30%.
  • • Integrated external APIs to improve data pipeline efficiency by 25%.

Technical Stack

AI & LLM

Prompt Engineering LangChain RAG Pipelines LLM Orchestration Multimodal AI WebGPU Local Inference Ollama AI Evaluation Agentic Workflows

Development

Python TypeScript JavaScript Next.js React Flask Django Streamlit SQL

Cloud & Infra

AWS (EC2, S3, CloudFront) Docker Vercel HLS Streaming FFmpeg

Product Strategy

Product Strategy Service Design UX & User Journey Design Business Process Automation Solution Architecture Digital Product Development Lead Generation

Analytics

Google Analytics 4 (GA4) User Behaviour Analytics Conversion Tracking SEO Fundamentals Performance Monitoring

Tools & Others

Git Postman Linux Tableau Excel Tesseract.js Telegram Bot API

Education

Master of Computer Applications (MCA)

Asansol Engineering College

2021 – 2023

CGPA: 8.93

Bachelor of Computer Applications (BCA)

Asansol Engineering College

2018 – 2021

CGPA: 8.59

Data Science & Machine Learning Specialization

Scaler Academy

2023 – 2025

Let's build
something epic.

Always open to discussing AI systems, automation, and collaborative innovation.

Email

ayanjawaid.mca2023@gmail.com