Skip to content
View yorayriniwnl's full-sized avatar

Highlights

  • Pro

Block or report yorayriniwnl

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.

Content in all repositories owned by your account will be closed.
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
Yorayriniwnl/README.md

Ayush Roy, Associate Engineer · Automotive Technology Services and Applied ML Builder

Jump to selected projects Jump to experience and education Jump to GitHub activity and profile views Jump to contact and collaboration

Currently Associate Engineer · Automotive Technology Services. I build ambitious product systems where polished interfaces meet realtime backends, applied machine learning, and the physical world. Open to jobs. Bengaluru (Bangalore), Karnataka, India.

Open Ayush Roy portfolio View Ayush Roy public resume Connect on LinkedIn Email Ayush Roy

Product, realtime systems, computer vision, and 3D interfaces

5 END-TO-END APPS. Designed and built across product, backend, realtime, and ML layers 24 AUTOMATED TESTS. Across five Vitest suites in the portfolio
78.5% HELD-OUT ACCURACY. Deterministic 80/20 holdout using LBP and GLCM features with an RBF SVM 3 DEVPOST PROTOTYPES. Shipped under competitive hackathon deadlines

Section 01: six selected projects

Jump to PORTFOLIO Jump to AI VS. REAL Jump to ZENITH Jump to HELIOS Jump to TOKEN USAGE Jump to TALKS V2

Personal Developer Portfolio: the product universe. Illustrative artwork.

Personal Developer Portfolio: Live. A focused portfolio with a lazy-loaded case-study architecture and a GPU-accelerated 4,000-point interface. Proof: 4,000 GPU-driven particles; 24 tests across five Vitest suites; Automated GitHub content synchronization. Stack: Next.js 16; TypeScript; Three.js; React Three Fiber; Framer Motion; Vitest.

Expand Personal Developer Portfolio project details

Personal Developer Portfolio: Live. A focused portfolio with a lazy-loaded case-study architecture and a GPU-accelerated 4,000-point interface. Proof: 4,000 GPU-driven particles; 24 tests across five Vitest suites; Automated GitHub content synchronization. Stack: Next.js 16; TypeScript; Three.js; React Three Fiber; Framer Motion; Vitest.

Open Personal Developer Portfolio project Inspect Personal Developer Portfolio source repository

AI vs. Real Image Detector: texture forensics. Illustrative artwork.

AI vs. Real Image Detector: Live. A probability-aware image classifier built from handcrafted texture features and a calibrated support vector machine. Proof: 78.5% held-out test accuracy; Local inference with committed SVM and scaler artifacts; Lighting-invariant preprocessing. Stack: Python; OpenCV; Scikit-Learn; LBP; GLCM; SVM; Streamlit.

Expand AI vs. Real Image Detector project details

AI vs. Real Image Detector: Live. A probability-aware image classifier built from handcrafted texture features and a calibrated support vector machine. Proof: 78.5% held-out test accuracy; Local inference with committed SVM and scaler artifacts; Lighting-invariant preprocessing. Stack: Python; OpenCV; Scikit-Learn; LBP; GLCM; SVM; Streamlit.

Open AI vs. Real Image Detector project Inspect AI vs. Real Image Detector source repository

Yor Zenith: solar decision support. Illustrative artwork.

Yor Zenith: Live. A full-stack solar feasibility platform combining 3D roof planning, energy-output simulation, and financial analysis. Proof: Interactive panel-placement simulation; Energy-output modeling across roof configurations; Subsidy, ROI, and payback analysis. Stack: React; TypeScript; Three.js; Python; FastAPI.

Expand Yor Zenith project details

Yor Zenith: Live. A full-stack solar feasibility platform combining 3D roof planning, energy-output simulation, and financial analysis. Proof: Interactive panel-placement simulation; Energy-output modeling across roof configurations; Subsidy, ROI, and payback analysis. Stack: React; TypeScript; Three.js; Python; FastAPI.

Open Yor Zenith project Inspect Yor Zenith source repository

Yor Helios: realtime energy intelligence. Illustrative artwork.

Yor Helios: In development. An energy-monitoring platform streaming anomaly and threshold events into channel-specific operator dashboards. Proof: Bidirectional realtime alert streaming; Continuous anomaly and threshold engine; One-command Docker Compose environment. Stack: Python; FastAPI; TypeScript; Docker; WebSocket.

Expand Yor Helios project details

Yor Helios: In development. An energy-monitoring platform streaming anomaly and threshold events into channel-specific operator dashboards. Proof: Bidirectional realtime alert streaming; Continuous anomaly and threshold engine; One-command Docker Compose environment. Stack: Python; FastAPI; TypeScript; Docker; WebSocket.

Inspect Yor Helios source repository

Yor_Token_Usage: multi-ai usage cockpit. Illustrative artwork.

Yor_Token_Usage: Experimental. A local-first Manifest V3 Chrome extension for estimating AI token usage across ChatGPT, Claude, Gemini, Perplexity, and Grok, with dashboards, exports, and optional sync. Proof: Local-first usage capture across five AI surfaces; Dashboards and exports for usage inspection; Optional sync with explicit privacy and clear-data controls. Stack: JavaScript; Manifest V3; Chrome Extensions; Local-first storage.

Expand Yor_Token_Usage project details

Yor_Token_Usage: Experimental. A local-first Manifest V3 Chrome extension for estimating AI token usage across ChatGPT, Claude, Gemini, Perplexity, and Grok, with dashboards, exports, and optional sync. Proof: Local-first usage capture across five AI surfaces; Dashboards and exports for usage inspection; Optional sync with explicit privacy and clear-data controls. Stack: JavaScript; Manifest V3; Chrome Extensions; Local-first storage.

Inspect Yor_Token_Usage source repository

Yor Talks V2: realtime communication. Illustrative artwork.

Yor Talks V2: Demo. A full-stack communication demo with realtime messaging, typed interface components, and a documented public-beta path. Proof: Low-latency bidirectional messaging; Authentication, conversation, and message APIs; Responsive reusable interface system. Stack: React; Vite; TypeScript; Express; Socket.IO; Postgres; Drizzle.

Expand Yor Talks V2 project details

Yor Talks V2: Demo. A full-stack communication demo with realtime messaging, typed interface components, and a documented public-beta path. Proof: Low-latency bidirectional messaging; Authentication, conversation, and message APIs; Responsive reusable interface system. Stack: React; Vite; TypeScript; Express; Socket.IO; Postgres; Drizzle.

Open Yor Talks V2 project Inspect Yor Talks V2 source repository

Section 02: experience and education

Telecom & Data Network Intern at Bharat Sanchar Nigam Limited (BSNL), June 2026. Completed an RGMTTC-certified four-week program in telecom, data-network systems, and infrastructure. B.Tech in Computer Science and Communication Engineering at KIIT Deemed University, 2023-2027 (Expected). Coursework: Data Structures & Algorithms; Operating Systems; Database Management Systems; Computer Networks; Object-Oriented Programming.

View the public resume

Section 03: technical range

Across the stack, from interface to infrastructure

product: TypeScript, JavaScript, React, Next.js, Tailwind CSS, Three.js, React Three Fiber, Framer Motion; backend: Python, FastAPI, Flask, Node.js, REST APIs, WebSocket, SQL, SQLite; ml: OpenCV, Scikit-Learn, SVM, LBP, GLCM, Feature Engineering, Model Evaluation, Streamlit; platform: Docker, Docker Compose, Git, GitHub Actions, Vercel, Linux, Vitest; expanding: Generative AI, LLMs, Prompt Engineering, RAG, AI Agents, LangChain, AWS S3, AWS Lambda, Vector Databases

Section 04: public GitHub activity

Total profile views, as counted by Komarev

GitHub repositories, stars, followers, and languages. Timestamped public snapshot.

365 days of GitHub contributions, with date range and activity totals

Section 05: the process and the person

Expand the field manual: process, principles, and off the clock Six project domains in motion. Illustrative animation, not live telemetry.

Achievements beyond the code

Engineering process

Product principles

Off the clock: visit Ayush Roy on Steam

Section 06: jobs and collaboration

Start a conversation with Ayush Roy

Email Ayush Roy Connect on LinkedIn Open the portfolio Follow on GitHub Explore Devpost prototypes Open Steam profile

Pinned Loading

  1. Yor-Ayrin-iwnl Yor-Ayrin-iwnl Public

    Production-grade 3D developer portfolio with 4,000 GPU-driven particles, lazy-loaded case studies, and 24 automated tests.

    CSS

  2. Yor-Ai-vs-real-image Yor-Ai-vs-real-image Public

    AI-vs-real image detector using LBP/GLCM texture features and a calibrated SVM — 78.5% held-out accuracy with local inference.

    Python

  3. Yor-Zenith Yor-Zenith Public

    Full-stack solar feasibility platform with 3D roof planning, energy simulation, subsidy, ROI, and payback analysis.

    TypeScript

  4. Yor-Helios Yor-Helios Public

    Realtime energy intelligence platform streaming anomaly and threshold alerts through FastAPI, WebSockets, and Docker.

    Python

  5. Yor_Token_Usage Yor_Token_Usage Public

    Local-first Manifest V3 Chrome extension for estimating AI token usage across ChatGPT, Claude, Gemini, Perplexity, and Grok, with dashboards, exports, and optional sync.

    TypeScript

  6. yor-talksv2 yor-talksv2 Public

    Full-stack realtime social-product platform with React, Vite, Express, Socket.IO, PostgreSQL, Drizzle ORM, messaging, creator tools, live, and shop surfaces.

    TypeScript 2 1