๐Ÿ“ Surat, India ๐ŸŽ“ B.E. Computer Engineering โ€” SSASIT, Surat โšก AI Product Architect ๐Ÿค Open to freelance, internships, open-source work, and serious product ideas Always Learning ;)

Hey, I'm

Roshan

AI Product Architect & Creative Technologist

I build AI products, automation systems, creative AI experiences, and marketing infrastructure.

From visual learning platforms and AI-powered apps to content engines, orchestration workflows, and business automation systems, I enjoy turning ideas into products people can actually use.

Currently building Diagrammatics, an AI-native visual learning platform, while experimenting with AI agents, cognitive interfaces, creative tooling, and automation systems that reduce friction for both businesses and individuals.

Computer Engineering student at SSASIT, Surat ยท Freelancer ยท Builder ยท Constantly exploring new tools, systems, and ideas.

๐Ÿง  AI Products & Apps โš™๏ธ Automation Systems ๐ŸŽจ Creative AI Systems ๐Ÿ“ˆ Marketing Infrastructure

Why I Build Things

Most of my projects start with a simple question:

"Why is this harder than it should be?"

Sometimes it's a confusing engineering topic. Sometimes it's a repetitive workflow. Sometimes it's my own inability to stay focused.

That curiosity has led me to build everything from AI orchestration systems and educational tools to productivity experiments and creative AI applications. I'm currently studying Computer Engineering at SSASIT, Surat, while freelancing, exploring new technologies, and building products that solve problems I personally experience.

My biggest interest isn't automation itself. It's understanding how people think, learn, get distracted, and interact with technology.

That's why many of my projects combine AI, psychology, visual learning, and system design rather than focusing on software alone.

โ™Ÿ Hobby Playing Chess (1500 ELO)
๐Ÿง  Interested In Psychology, Cognition, Learning Systems
๐Ÿš€ Currently Building Diagrammatics
๐ŸŽฏ Current Goal Finish engineering while shipping real products
๐Ÿ”ฌ Exploring AI Agents, RAG, Cognitive Interfaces
๐Ÿ“ Based In Surat, Gujarat

Technical Skills

๐Ÿง  AI Systems & Product Engineering

AI Agents & WorkflowsLLMsRAGMulti-Agent SystemsAI Product ArchitecturePrompt EngineeringAI-Native UXVisual Learning

โš™๏ธ Workflow Orchestration

n8nAPI IntegrationsEvent-Driven WorkflowsWebhooksData PipelinesAutomation Design

๐Ÿ’ป Development

PythonTypeScriptJavaScriptSQLHTML5/CSS3FastAPINode.jsReactNext.js

โ˜๏ธ Infrastructure & Data

Gemini APIOpenAIAgentScopeVector DBsSupabaseGCPDockerVercel

๐ŸŽจ Product & Design

System DesignProduct ArchitectureHuman-Centered AIUX PsychologyVisual Info Design

The Schemata System

An AI-native visual learning platform designed to transform dense PDFs, STEM content, and technical concepts into diagrams, visual explanations, comparison grids, formula cards, and adaptive learning experiences. Focused on reducing cognitive load through visual-first understanding rather than information overload.

Visual Learning Local RAG AgentScope Memory Multimodal AI Mermaid FastAPI Next.js
View Live System →
The Schemata System Visual Interface Cognitive Visual UI Local RAG Core AgentScope Architecture

ADHD Project 01: Attention Redirect

Research & Ideation

A behavior-interruption system designed to detect distraction patterns, calm cognitive overload, and redirect users back to meaningful work through tiny actions and low-friction interventions.

ADHD Project 02: Executive Assistant for ADHD Minds

Research & Ideation

An autonomous AI companion designed to organize information from platforms such as WhatsApp and Google Classroom, reducing mental clutter and helping users stay focused on what matters.

Social Speaking → Gemini + React
SpeechRhythm

SpeechRhythm

"Duolingo for spontaneous speaking." An intensive speaking challenge engine featuring 1-minute prompts with camera + mic recording and Gemini scorecard (energy, clarity, hook, improvement points).

Gemini API Tailwind Framer Motion
Neuro-Focus → Kotlin
ADHDAssist

ADHDAssist

Mindful productivity companion for neurodivergent minds. Features distraction interception, thought vault, breathing anchor, and customizable low-stimulation co-regulation modes. Calm sketchbook style UI.

Compose Room DB Gemini
EdTech Assistant → TS + React
GTU Copilot

GTU Copilot

High-fidelity academic assistant for GTU students. Features GTU Expert mode, LaTeX rendering, visual aids, custom document generation, reports, and slide outputs.

Tailwind pptxgenjs Gemini
Design Thinking → React
SSASIT Mapping

SSASIT Mapping

Design Thinking campus navigation system. Displays interactive digital maps to help parents, students, and visitors navigate the SSASIT campus seamlessly.

Vite Zoom-Pan-Pinch Motion
Academic Platform → React Flow
AI Assignment Solver Studio

AI Assignment Solver Studio

AI-powered academic workflow platform generating editable diagrams, viva questions, and exam answers. Features split workspace layout, section regeneration, custom handwriting styles, and AI thinking timeline.

Zustand React Flow Gemini API
Astrology UI → React
Cosmic.AI

Cosmic.AI

Minimalist astrology and decision compass app powered by Gemini. Features glassmorphism design with color decision alerts (Green / Yellow / Red).

Tailwind Lucide Icons Gemini
AI Studio → Live
NanoMorph

NanoMorph

AI-powered photo transformation studio for instant edits and enhancements. Deployed inside Google AI Studio.

Gemini Flash Image API
AI Studio → Live
Visual Chapter Plan

Visual Chapter Plan

Editorial tool to transform narrative texts into structured visual plans and scene-by-scene imagery.

Gemini Flash Asset Generation
AI Studio → Live
Lumina Product Studio

Lumina Product Studio

AI e-commerce asset generator creating backgrounds and mockups directly from product snapshots.

Gemini Flash Product Mockups
AI Studio → Live
Lumina Creative Canvas

Lumina Creative Canvas

Node-based creative workspace to chain AI images, edits, prompts, and visual models in an interactive workspace interface.

Gemini Flash Infinite Canvas AI Node Pipelines

I use AI-assisted filmmaking and visual generation workflows to explore storytelling, marketing, and creative system design.

The Whispering Cloud

Short Film / Mythic Theme
๐ŸŽฌ Whispering Cloud
๐ŸŽฎ PocketMC
๐Ÿช Antigravity
๐Ÿง Evolution of Linux
โ™Ÿ๏ธ Chess.com
Enterprise B2B Sales Orchestrator

Enterprise B2B Sales Orchestrator

50+ node production-grade sales funnel. Automates lead extraction (Whisper), dynamic multi-step cold email sequencing (Mailgun), and automated lead tracking (Supabase).

n8n OpenAI Whisper Mailgun Supabase CRM Integration
Reels Automate: Viral AI News Agent

Steeve Automate: Viral AI News Agent

End-to-end AI video pipeline that monitors RSS news, scripts viral content with GPT-4o, generates AI avatars (HeyGen), and composites split-screen shorts (Creatomate) automatically.

n8n GPT-4o HeyGen Creatomate Postiz
Meta Ad Campaign Automation

Meta Ad Campaign Automation

54-node orchestration workflow automating campaign creation, creative rotation, and multi-platform budget management.

n8n Meta Ads API Automation
Instagram Auto-Posting System

Instagram Auto-Posting System

35-node workflow handling scheduling, caption generation, hashtag suggestion, and image optimizations via editorial calendar.

n8n Instagram API Image Tooling
Document-Based Chatbot

Document-Based Chatbot (Q&A)

Ingest document corpus โ†’ vectorize โ†’ retrieval โ†’ LLM summarization chat UI. Reduces lookup and support time significantly.

Embeddings FAISS n8n Chat UI

๐Ÿš€ Custom AI Orchestration & System Design

Hire me to modify these workflows, setup custom n8n systems, build your app/startup, or architect your technical infrastructure.

Gamma Watermark Remover

Gamma Watermark Remover

Local-first tool to remove the "Made with Gamma" watermark from PDF and PPTX exports. Studio UI with batch processing and privacy focus.

Python PDF PPTX MIT
GTU Design Thinking Canvas

GTU Design Thinking Canvas

High-fidelity interactive digital workspace targeted at GTU Design Engineering curriculum with standardized canvases.

TypeScript Canvas API WebApp
SemesterScan

SemesterScan

Privacy-first AI organizer for college WhatsApp groups. Automatically categorizes and organizes academic content while respecting user privacy.

JavaScript AI Privacy MIT
Jyotish.AI

Jyotish.AI

Minimalist Vedic astrology and numerology app powered by Google Gemini for clear, actionable insights.

TypeScript Gemini AI Astrology
Scholar Claw / StudyClaw

Scholar Claw (StudyClaw)

AI-native study infrastructure integrating Gemini AI into syllabuses via WhatsApp/Telegram. Ingests curriculum into structured chat flows.

Gemini AI WhatsApp Bot Academic AI
๐Ÿ…

Develop GenAI Apps with Gemini and Streamlit

Nov 8, 2025

๐Ÿ…

Prompt Design in Vertex AI

Nov 8, 2025

๐Ÿ…

Analyze Speech & Language with Google APIs

Nov 7, 2025

๐Ÿ…

Intro to ML: Language Processing

2025

๐Ÿ…

Google Cloud Computing Foundations

2025

๐Ÿ…

Build & Deploy ML Solutions on Vertex AI

2025

๐Ÿ…

Generative AI Fundamentals

2025

๐Ÿ…

Responsible AI Practices

2025

๐Ÿ…

Cloud Architecture Design

2025

LET'S CONNECT

Most of the things on this website started as random ideas, frustrations, side experiments, or questions I couldn't stop thinking about.

If you're building something interesting in AI, education, productivity, automation, design, or emerging technology, I'd be happy to connect.

I'm currently open to:

  • โ€ข Internship opportunities
  • โ€ข Startup collaborations
  • โ€ข Open-source projects
  • โ€ข Student and research initiatives
  • โ€ข Product discussions
  • โ€ข Interesting ideas worth exploring

Whether it's a project, opportunity, feedback, or simply a conversation about technology, feel free to reach out.

Currently based in Surat, Gujarat.

Let's build something useful.