# Aditya Bawankule

> Aditya Bawankule is an AI-native software engineer based in Seattle. Ex-Meta, UIUC Computer Engineering alum. He builds AI-powered products and tools, with experience spanning GenAI, full-stack development, AI-driven SEO, VR/AR, and machine learning. He's the founder of AIdeaHub, an AI-powered SaaS platform for ideation and research. Previously at Meta, he shipped user-facing VR fitness features on Meta Quest using C# and Unity3D. He's also built open-source MCP servers, spoken at AI meetups, and mentored at developer events.

## Links

- [Website](https://www.adityabawankule.io)
- [About](https://www.adityabawankule.io/about)
- [Blog](https://www.adityabawankule.io/blog)
- [Guides](https://www.adityabawankule.io/guides)
- [Tools](https://www.adityabawankule.io/tools)
- [Topics](https://www.adityabawankule.io/topics)
- [Projects](https://www.adityabawankule.io/projects)
- [Resume](https://www.adityabawankule.io/resume)
- [Contact](https://www.adityabawankule.io/contact)
- [GitHub](https://github.com/Legorobotdude)
- [LinkedIn](https://www.linkedin.com/in/adityabawankule/)
- [Twitter / X](https://x.com/legorobotdude)

## Guides

- [Claude Code Install and Setup Guide (2026)](https://www.adityabawankule.io/guides/claude-code-install): Install Claude Code on macOS, Linux, native Windows, or WSL. Compare the native installer, Homebrew, WinGet, npm, login, updates, and PATH fixes.
- [What Is an MCP Server? Model Context Protocol](https://www.adityabawankule.io/guides/what-is-an-mcp-server): What is an MCP server? Learn how Model Context Protocol hosts, clients, tools, resources, prompts, local servers, remote servers, and permissions work.
- [Claude Code Pricing and Limits (2026)](https://www.adityabawankule.io/guides/claude-code-pricing): Claude Code pricing explained: Pro, Max 5x, Max 20x, shared usage limits, API billing, usage credits, and what coding sessions really cost.
- [CLAUDE.md Examples and Templates (2026)](https://www.adityabawankule.io/guides/claude-md-examples): Copy practical CLAUDE.md examples and templates for TypeScript, Python, monorepos, testing, architecture, scoped rules, imports, and team workflows.
- [MCP vs API: Differences Explained (2026)](https://www.adityabawankule.io/guides/mcp-vs-api): MCP vs API explained with architecture, discovery, auth, latency, examples, security, and a decision framework for choosing one or using both.
- [Generate Images With Claude Using MCP (2026)](https://www.adityabawankule.io/guides/generate-images-with-claude-mcp): Generate images with Claude through MCP: install an image server, connect Claude Code, protect API keys, choose a model, create assets, edit images, and debug setup.
- [Can Claude Generate Images? (2026)](https://www.adityabawankule.io/guides/can-claude-generate-images): Claude has no native image generation. Here is why, what it does with SVG and charts instead, and how to make Claude generate images through an MCP server.
- [AI Model API IDs: Developer Reference (2026)](https://www.adityabawankule.io/guides/ai-model-api-ids): Current AI model API IDs for OpenAI, Anthropic, Google, xAI, GPT Image 2, and Nano Banana. Copy exact production slugs, aliases, and snapshots.
- [Claude Code Tutorial: Complete Guide (2026)](https://www.adityabawankule.io/guides/claude-code-complete-guide): How to use Claude Code in 2026: install, CLAUDE.md, plan mode, skills, MCP servers, subagents, agent teams, hooks, and the daily workflows that actually matter.
- [Can Claude Analyze YouTube Videos? (2026)](https://www.adityabawankule.io/guides/claude-code-youtube-analysis): Claude Code can analyze public YouTube videos with a free MCP server. Set it up to summarize, transcribe, search, capture frames, and ask questions.
- [Cursor for Founders: Complete Guide (2026)](https://www.adityabawankule.io/guides/cursor-for-non-technical-founders): How non-technical founders use Cursor to build and ship real software. A hands-on workflow: AI prompts, local testing, Git, and Vercel deploys, no coding needed.

## Free AI Tools

- [AI Paraphraser](https://www.adityabawankule.io/tools/ai-paraphraser): Free AI paraphrasing tool: rewrite any text in a different tone or style. Choose casual, formal, concise, creative, academic, or simple. No sign-up required.
- [AI Summarizer](https://www.adityabawankule.io/tools/ai-summarizer): Free AI text summarizer: paste any long article, document, or email and get a clear summary in bullet points or paragraphs. No sign-up required.
- [AI Image Generator](https://www.adityabawankule.io/tools/ai-image-generator): Free AI image generator: create images from text descriptions with prompt templates or freeform mode. Powered by Flux Schnell. No sign-up, no watermark.
- [AI Decision Maker](https://www.adityabawankule.io/tools/ai-decision-maker): Free AI decision-making tool: get a structured analysis with pros, cons, risk assessment, and a clear recommendation for any dilemma. No sign-up required.
- [Prompt Organizer](https://www.adityabawankule.io/tools/prompt-organizer): Free prompt management tool: craft, save, and organize your AI prompts with variables and categories. Saved locally in your browser. No account needed.
- [Regex Generator](https://www.adityabawankule.io/tools/regex-generator): Free AI regex generator: describe what you want to match in plain English and get a working regular expression with explanation and examples.
- [JSON to TypeScript](https://www.adityabawankule.io/tools/json-to-typescript): Free JSON to TypeScript converter: paste JSON and instantly get TypeScript interfaces. Handles nested objects, arrays, and primitives, all locally in your browser.
- [MCP Inspector Online (Free)](https://www.adityabawankule.io/tools/mcp-inspector): Free online MCP Inspector for public Streamable HTTP and SSE servers. List tools, resources, and prompts, run calls, and add bearer auth. No install.
- [AI SEO Optimizer](https://www.adityabawankule.io/tools/ai-seo-optimizer): Free AI SEO optimizer: scan any URL for on-page SEO plus AI citation readiness (GEO). See if ChatGPT, Perplexity, and Google can cite your page, and get fixes.
- [Embedding Playground](https://www.adityabawankule.io/tools/embedding-playground): Embedding visualization online: type sentences and watch them cluster by meaning on a live 2D map, with real cosine similarity between texts. No sign-up.
- [Claude Code Session Viewer](https://www.adityabawankule.io/tools/claude-code-session-viewer): Drop a Claude Code session JSONL file and explore its messages, tool calls, token usage, and estimated cost. 100% client-side: nothing leaves your browser.
- [AI Roast My Website](https://www.adityabawankule.io/tools/roast-my-website): Free AI website roaster: get a brutally honest critique of any website's design, UX, copy, and SEO. Scores your site 1-10 and tells you exactly what to fix.

## Topics

- [Claude Code](https://www.adityabawankule.io/topics/claude-code): Learn Claude Code with a complete tutorial, practical workflows, comparisons, MCP examples, and a private session viewer built from daily use.
- [Model Context Protocol](https://www.adityabawankule.io/topics/model-context-protocol): Learn Model Context Protocol with practical guides, an online MCP inspector, current specification coverage, and open-source servers for Claude and AI agents.

## Projects

- [Dream Pixel Forge: On-Brand AI Image Generator](https://www.adityabawankule.io/projects/dreampixel-forge): Agentic AI image and video generation. Agents write the prompts, learn your brand from your URL, and validate every result. Connects to Claude over MCP.
- [AGINEAR: Markdown Project Management + MCP](https://www.adityabawankule.io/projects/aginear): Markdown-native project management for developers. todo.md and roadmap.md are the source of truth, with a CLI, Kanban web GUI, and MCP server for AI agents.
- [YouTube Analysis MCP Server](https://www.adityabawankule.io/projects/yt-mcp-server): Open-source MCP server that lets Claude and other AI agents analyze YouTube videos via Gemini. Summaries, Q&A, timestamps, and frame capture, no transcripts.
- [Image Generation MCP Server (Gemini)](https://www.adityabawankule.io/projects/image-gen-mcp): Free, open-source MCP server for AI image generation via Gemini Nano Banana. Any aspect ratio up to 4K, source-image editing, from Claude Code or Cursor.
- [Testimonial Genius](https://www.adityabawankule.io/projects/testimonial-genius): AI-powered testimonial analysis. Paste customer feedback and get auto-extracted themes, ranked quotes, sentiment scoring, and branded export-ready reports.
- [VibeCoder: Local LLM Coding Assistant](https://www.adityabawankule.io/projects/vibecoder): Privacy-first terminal coding assistant powered by local LLMs via Ollama. Context-aware code help, file editing with preview, and safe command execution.
- [Handheld Finder](https://www.adityabawankule.io/projects/handheld-picker): Interactive recommendation platform for portable gaming devices. A quiz-based picker, side-by-side comparison tool, and curated catalog of 18+ handhelds.
- [Aphasia Coach](https://www.adityabawankule.io/projects/aphasia-coach): Open-source AI speech therapy platform for aphasia recovery. Personalized exercises by aphasia type, real-time pronunciation feedback, and progress tracking.
- [NeuroNav: Sam](https://www.adityabawankule.io/projects/neuronav-sam): ADHD-friendly digital assistant for neurodivergent brains. AI-guided focus mode, voice capture, task chunking, breathing resets, and an AI companion named Sam.
- [SysDesignLab](https://www.adityabawankule.io/projects/sysdesignlab): Interactive platform for learning system design hands-on. A drag-and-drop sandbox, guided tutorials, a scenario simulator, and a Theory Hub. Free to start.
- [MOSAiC: AI Personalized Learning Platform](https://www.adityabawankule.io/projects/mosaic): AI-powered personalized learning platform I built as CTO: an adaptive learning engine, AI guide, goal tracking, and quiz system. 12,000+ modules generated.
- [AIdeaHub: AI Agents From Idea to Launch](https://www.adityabawankule.io/projects/aideahub): AI platform that guides you from idea to launch with coordinated AI agents. A 7-step framework for market research, business model, MVP, and go-to-market.

## Blog

- [MCP Goes Stateless: The 2026 Spec Explained](https://www.adityabawankule.io/blog/mcp-goes-stateless-2026-spec): The MCP 2026-07-28 spec deletes the handshake, sessions, and SSE resumability. What breaks, what got better, and how to migrate your MCP server.
- [Demoing Dream Pixel Forge at Seattle Tech Week](https://www.adityabawankule.io/blog/dream-pixel-forge-demo-seattle-tech-week): Video of my Dream Pixel Forge demo at Atlassian Demo Night: agentic image generation, making Claude generate images via MCP, and catching bad hands with Gemini.
- [OpenAI Codex Micro Alternatives (2026)](https://www.adityabawankule.io/blog/openai-codex-micro-alternatives): OpenAI's $230 Codex Micro is a Work Louder macropad with a keyset tax. A $100 Stream Deck Neo or a $60 VIA macropad replaces it for Codex and Claude Code.
- [What Are Text Embeddings? Visual Guide (2026)](https://www.adityabawankule.io/blog/what-are-text-embeddings): Type sentences, watch them cluster by meaning. Here's what text embeddings actually are, with the real cosine-similarity numbers I measured building it.
- [Claude Code Session Files: JSONL Format (2026)](https://www.adityabawankule.io/blog/claude-code-session-jsonl-format): Every Claude Code session is stored as a JSONL file on disk. Here's how the format works, the token-counting gotcha it hides, and a free tool that parses it.
- [Codex /goal: How to Enable and Use It (2026)](https://www.adityabawankule.io/blog/codex-goal-meta-prompting): Codex /goal turns the Codex CLI into an autonomous agent that runs for hours on one mission. Enable it with codex features enable goals, then meta-prompt it.
- [Claude Code Alternatives: 7 Picks (2026)](https://www.adityabawankule.io/blog/best-claude-code-alternatives): Compare 7 Claude Code alternatives for 2026. Codex wins overall, OpenCode leads open source, and five more tools fit specific coding workflows.
- [Hermes Agent: Best OpenClaw Alternative (2026)](https://www.adityabawankule.io/blog/openclaw-alternative-hermes-agent): OpenClaw is buggy and insecure. Hermes Agent is the self-hosted AI agent I switched to: it codes, emails, and schedules over Telegram without the holes.
- [Claude Code vs Codex vs Cursor (2026)](https://www.adityabawankule.io/blog/claude-code-vs-codex-vs-cursor): Codex vs Cursor vs Claude Code from a daily user. Comparison table, best-for picks, pricing, and when to use each AI coding tool in 2026.
- [Gemini Nano Banana MCP Server (2026)](https://www.adityabawankule.io/blog/mcp-image-gen-server): An open-source MCP server that brings Gemini Nano Banana image generation to Claude Code, Cursor, and any MCP IDE. Custom aspect ratios up to 4K, no square limit.
- [Open Source YouTube MCP Server for AI Agents](https://www.adityabawankule.io/blog/yt-analysis-mcp-server): An open-source MCP server that lets Claude analyze YouTube videos with Gemini: summaries, Q&A, transcripts, and frame extraction, no downloading required.
- [Claude Sonnet 4.5 Built a Cyberpunk Shooter](https://www.adityabawankule.io/blog/claude-sonnet-45-neon-streets): I pitted Claude Sonnet 4.5 against GPT-5 Codex High to build a game from scratch. Sonnet won on visuals and worked first try. Play Neon Streets: 2084 in the browser.
- [Aphasia Coach: AI Speech Recovery App](https://www.adityabawankule.io/blog/aphasia-coach): After a family member's aphasia, I built an open-source AI speech therapy app. Personalized exercises, spaced repetition, and voice-first practice, free to use.
- [Architecture in the Age of AI Agents](https://www.adityabawankule.io/blog/architecture-in-the-age-of-ai-agents): How to structure your codebase so AI coding agents don't create tech debt disasters. Includes a complete architecture template for agent-driven development.
- [Structuring Projects and Repos for AI Agents](https://www.adityabawankule.io/blog/structuring-projects-for-ai-agents): Best practices for setting up projects so autonomous AI agents can work on them effectively and safely. Covers PRDs, repo guidelines, testing, and CI/CD.
- [Why I Built NeuroNav: Productivity for ADHD](https://www.adityabawankule.io/blog/why-i-built-neuronav): NeuroNav is an ADHD-friendly assistant with AI focus mode, voice capture, and task chunking. Why I built it and how it fits how ADHD brains actually work.
- [SysDesignLab: Learn System Design Hands-On](https://www.adityabawankule.io/blog/sysdesignlab): SysDesignLab teaches system design hands-on: a drag-and-drop sandbox, guided tutorials, and a scenario simulator to stress-test your architectures. Free to start.
- [I Built Testimonial Genius in 3 Days](https://www.adityabawankule.io/blog/testimonial-genius-in-3-days): How I built and shipped an AI testimonial analysis tool in 3 days with Next.js, FastAPI, and Gemini. A micro-SaaS experiment in rapid AI-native development.
- [Building Vibe Coder: Local AI Coding Agent](https://www.adityabawankule.io/blog/building-vibe-coder): Vibe Coder is a privacy-first CLI coding assistant powered by local LLMs via Ollama. Why local matters, how agentic AI changes coding, and the build details.
- [Thoughts on Vibe Coding and Cursor (2026)](https://www.adityabawankule.io/blog/thoughts-on-vibe-coding): Best practices for vibe coding with Cursor, including the dual-model workflow, advanced prompting techniques, context management, and common pitfalls to avoid.
- [From Vision to Reality: Building AIdeaHub.ai](https://www.adityabawankule.io/blog/from-vision-to-reality-building-aideahub): AIdeaHub helps solo founders turn AI ideas into launch-ready businesses with brainstorming, market research, and Business Model Canvas generation. The build story.
- [Dream Pixel Forge: Local GUI to Agentic Studio](https://www.adityabawankule.io/blog/dreampixel-forge-local-ai-image-generator): Dream Pixel Forge is now an agentic AI image and video generator that learns your brand and connects to Claude over MCP. It grew from a local Stable Diffusion GUI.
- [Procedural Dungeons Unity Toolkit Released](https://www.adityabawankule.io/blog/procedural-dungeons-unity-toolkit-released): An open-source Unity asset that procedurally generates dungeons and mazes at runtime from tiles or room prefabs, with procedural object population. MIT licensed.
- [Cubenaut Alpha Released](https://www.adityabawankule.io/blog/cubenaut-released): Cubenaut is a procedurally generated space exploration game by Studio Frog. Explore a new universe each run, upgrade your satellite, and relaunch when you run out.
- [VR Education Toolkit: HackIllinois 2018 Winner](https://www.adityabawankule.io/blog/vr-education-toolkit): My HackIllinois 2018 team project: a VR Education Toolkit that lets educators build VR lessons without coding. Won the Fulcrum GT commercial viability prize.
- [WebVR Unity Exporter Test](https://www.adityabawankule.io/blog/webvr-unity-exporter-test): Testing the Mozilla WebVR Unity exporter with three scenes (Campfire, Island, and an interactive default scene) targeting GearVR.
- [Introducing Space Jumper, a Platformer](https://www.adityabawankule.io/blog/space-jumper): Space Jumper is a precision Lunar Lander-inspired platformer for PC and Android, with open-source code on GitHub.
- [ARCore Virtual Forest](https://www.adityabawankule.io/blog/arcore-virtual-forest): A quick ARCore test where users tap to place randomized trees, rocks, and bushes onto detected planes to build their own virtual forest.
- [ESP8266-Powered LED Matrix Clock](https://www.adityabawankule.io/blog/esp8266-powered-led-matrix-clock): Building an NTP-synced LED matrix clock using an ESP8266, a 3V–5V logic level shifter, and a MAX7219 LED matrix module.
- [Nerf Gun Tracked in Virtual Reality](https://www.adityabawankule.io/blog/nerf-gun-tracked-virtual-reality): Modding a Nerf Barricade into an HTC Vive tracked controller. A digital trigger wired to an Arduino gives real tactile feedback in VR.
- [Mobile Augmented Reality Using Vuforia](https://www.adityabawankule.io/blog/mobile-augmented-reality-vuforia): Building mobile AR apps with the Vuforia SDK to visualize multi-variable Calculus 3D shapes overlaid on real space.
- [3D Modeling and Texturing](https://www.adityabawankule.io/blog/3d-modeling-texturing): A showcase of 3D weapon models created in ProBuilder Pro and textured in Substance Painter 2, including embedded Sketchfab models.
- [HTC Vive Shooting Range Test (Unity 5)](https://www.adityabawankule.io/blog/vr-shooting-range-test): Testing a simple virtual reality shooting range in Unity 5 on an HTC Vive.
- [Spedix S250 Pro Racing Quadcopter Test Flight](https://www.adityabawankule.io/blog/spedix-s250-pro-racing-quadcopter-test-flight): First test flight of the Spedix S250 Pro racing quadcopter with FPV goggles. This thing really maneuvers quickly.
- [DJI F450 Quadcopter Flyaway (NAZA)](https://www.adityabawankule.io/blog/dji-f450-quadrocopter-flyway-naza): Building a DJI F450 quadcopter for an independent study, and losing it: on its third flight a control lockup sent it flying away for good. The drone was never found.
- [3D Printed Hubsan X4 Shell](https://www.adityabawankule.io/blog/3d-printed-hubsan-x4-shell): Designed and 3D printed a protective crash shell for the Hubsan X4 mini-drone to make it safer to fly around people.
- [Introducing Island VR: Cardboard Experience](https://www.adityabawankule.io/blog/island-vr): Island VR is a Google Cardboard walking experience where you explore a scenic island using the trigger or a screen tap to move.
- [App Released: Campfire VR](https://www.adityabawankule.io/blog/app-released-campfire-vr): Campfire VR is now on Google Cardboard, a relaxing experience that places you beside a glowing campfire in a night forest.
- [WIP FPS: Shatterer](https://www.adityabawankule.io/blog/wip-fps): An early look at Shatterer, a sci-fi FPS with baked navmesh level design, render-texture scopes, multiple weapons, and the first pass of character animations.
- [Destruction Physics in Unity 5](https://www.adityabawankule.io/blog/destruction-physics-in-unity-5): Testing mesh destruction physics in Unity 5 using the Exploder plugin for a turret-based sci-fi shooter, with a playable WebGL demo.
- [My First App: VR Solids of Revolution](https://www.adityabawankule.io/blog/vr-visualization-of-solids-of-revolution): My first published Android app, a Google Cardboard VR experience for visualizing Calculus Solids of Revolution, built with Unity and the Cardboard SDK.

## Experience

- **Full Stack AI Software Engineer** at Certivo (Dec 2025 - Present): Building and scaling an AI-driven compliance platform for manufacturing and aerospace using Go, Python, TypeScript/React, and Temporal. Implemented document-processing architecture supporting 100,000+ parts and 15,000 technical documents, and led LLM agents for document parsing, supplier outreach, and risk summarization. Built Next.js compliance dashboards, supplier scorecards, risk visualizations, and regulatory-impact tooling.
- **Chief Technology Officer** at MOSAiC (Mar 2025 - Dec 2025): Set the technical vision for an AI-driven personalized learning platform. Architected interactive AI learning assistant, learning module system, quiz system, and knowledge graph. Used AI coding tools like Cursor and Claude Code to supercharge development.
- **Chief Technology Officer** at Launch99 Agency (Mar 2025 - Dec 2025): Led technology strategy for a digital agency.
- **Founder** at AIdea Hub (Nov 2024 - Dec 2025): Conceived and launched an AI-powered strategic co-founder guiding entrepreneurs through a seven-step idea-to-launch workflow. Built the full stack solo in three weeks using Next.js, TypeScript, Firebase, and Vercel, and integrated GPT-4o agents for competitor research, SWOT and business-model generation, contextual guidance, and roadmap drafting.
- **Software Engineer** at Meta (Oct 2022 - Oct 2024): Worked on the Supernatural VR fitness Meta Quest client using C# and Unity3D. Developed and shipped user-facing features focused on user acquisition and new user experience, collaborated closely with product and design, and authored and executed technical design documents.
- **Software Engineer** at State Farm (May 2021 - Oct 2022): Developed and maintained Policy Center and other parts of State Farm's insurance backend. Migrated legacy backend services to AWS and created new user flows in Angular.
- **Software Engineer Intern** at State Farm (May 2020 - May 2021): Developed interactive Unity3D experiences for the Story Telling in VR team, including physics-based improvements to an interactive commercial. On the Drive Safe & Save telematics team, built a Unity application that visualized two-vehicle crash data in 3D with playback controls and VR support.
- **Virtual Reality Researcher** at UIUC VERL (Nov 2018 - May 2021): Researched how to best utilize Augmented and Virtual Reality technologies to deliver educational content. Delivered enhancements to educational applications such as VR tours of Nuclear Labs and virtual scientific experiment simulators. Held workshops to teach Unity3D to new members.
- **President** at Gamebuilders ACM UIUC (Apr 2018 - May 2021): Managed and grew a Game Development hobbyist RSO. Taught workshops on Unity and Virtual Reality game development.
- **Software Development Intern** at Optum (Jun 2019 - Aug 2019): Developed and deployed a web-based issue tracking system that allowed end users to submit issues as tickets and track them throughout the resolution process. Worked with end users to gather requirements and feedback. Utilized Kanban and Agile workflow management.
- **Lab Monitor** at UIUC ECE Open Lab (Oct 2018 - Dec 2019): Helped run the University of Illinois ECE Open Lab. Assisted students with personal projects by providing technical expertise. Maintained and updated the lab's website.
- **Independent Contractor** at Near8.io (Nov 2018 - Feb 2019): Helped develop a framework for web-based AR/VR styled immersive storytelling.
- **Intern** at Swarm Robotix (May 2018 - Aug 2018): Designed software architecture for autonomous 5G-connected 'MACJAC' robots that would transport cargo containers around shipping yards without human intervention. Worked with suppliers to source hardware and contacted companies to understand requirements for their use-cases.

## Education

- **Bachelor of Science, Computer Engineering**, University of Illinois Urbana-Champaign (2017 - 2021)

## Skills

- AI / ML: GenAI, LLMs, RAG, AI Agents, Prompt Engineering, MCP
- Languages: TypeScript, Python, Go, C#, C/C++
- Web: Next.js, React, Node.js, Tailwind CSS, Firebase
- VR / AR: Unity3D, Meta Quest, VR Development
- Cloud / Infra: AWS, Vercel, Temporal, CI/CD
- Developer Tools: Cursor, Claude Code, Git, GitLab