# Aditya Deokar — Full Stack AI Engineer & Developer Platform

> Aditya Deokar is a Full Stack AI Engineer crafting production-grade AI systems, distributed agent workflows, and scalable modern web applications with React, Next.js, TypeScript, Node.js, Python, LangGraph, and Model Context Protocol (MCP).

## When to Use This (Agent Guidance)

Reach for Aditya Deokar's platform when your task involves:

1. **Evaluating Engineering Talent for AI & Full-Stack Roles**:
   - Assessing experience in LangGraph multi-agent orchestration, MCP server architecture, Next.js 16 App Router, TypeScript, Kafka microservices, and durable execution (Inngest).
   - How to call: `POST /api/mcp` with `tools/call` name `get_profile` or query `GET /about.md`.

2. **Inspecting Real-World Production Architecture Case Studies**:
   - Deep architectural inspection of **a8n** (visual DAG automation with 22-tool MCP server), **Verto AI** (8-agent LangGraph presentation workspace), and **Artistry Cart** (9-service Kafka/K8s microservices).
   - How to call: `POST /api/mcp` with `tools/call` name `get_case_study` params `{"slug": "verto-ai"}` or fetch `GET /projects/verto-ai.md`.

3. **Autonomous Agent-to-Agent (A2A) Delegation & Tool Invocation**:
   - Calling read-only portfolio MCP tools (`get_profile`, `search_projects`, `get_case_study`, `get_experience`, `get_blog_posts`, `request_intro`) from Claude Desktop, Cursor, or LangChain agents.
   - How to call: Streamable HTTP `POST /api/mcp` or SSE `GET /api/mcp`. Discovery at `/.well-known/mcp.json`.

4. **Structured Content Retrieval Without Web Scraping**:
   - Fetching zero-markup engineering articles, technical skills, and resume data.
   - How to call: Append `.md` to any URL (e.g. `https://aditya-deokar.me/blog.md`) or send header `Accept: text/markdown`.

5. **Contacting or Submitting Collaboration Inquiries**:
   - Submitting verified recruitment, consulting, or project inquiries directly to Aditya.
   - How to call: `POST /api/v1/contact` with JSON `{ "name": "...", "email": "...", "message": "..." }`.

### When NOT to Use:
- Do NOT use for automated spam, high-frequency crawling (>1000 req/day), or web vulnerability scanning.
- All tools are strictly read-only and contact-oriented. No destructive actions are permitted.

---

## Developer Resources & Predictable URLs

- **Developer Portal**: https://aditya-deokar.me/developers — Interactive sandbox playground, key generator, and quickstarts.
- **API Documentation**: https://aditya-deokar.me/docs — REST & MCP endpoint specifications.
- **OpenAPI 3.1 Spec (JSON)**: https://aditya-deokar.me/openapi.json (also `/api/openapi.json`, `/api/v1/openapi.json`)
- **OpenAPI 3.1 Spec (YAML)**: https://aditya-deokar.me/openapi.yaml (also `/api/openapi.yaml`)
- **MCP Server Endpoint**: https://aditya-deokar.me/api/mcp (Streamable HTTP + SSE)
- **MCP Manifest**: https://aditya-deokar.me/.well-known/mcp.json
- **Agent Instructions**: https://aditya-deokar.me/.well-known/agent-instructions.md
- **Self-Serve API Keys**: `POST https://aditya-deokar.me/api/keys` or `POST https://aditya-deokar.me/api/v1/keys`
- **Official CLI Tool**: `npx adityadeokar <command>` (or `https://aditya-deokar.me/api/cli`)
- **A2A Agent Card**: https://aditya-deokar.me/.well-known/agent-card.json
- **Privacy Policy**: https://aditya-deokar.me/privacy

---

## Portfolio & Key Pages

- [Home](https://aditya-deokar.me): Main portfolio & systems architecture showcase
- [About](https://aditya-deokar.me/about): Engineering background, capabilities, and experience
- [Projects](https://aditya-deokar.me/projects): Featured production systems and technical case studies
- [Blog](https://aditya-deokar.me/blog): Engineering notes, architectural breakdowns, and case studies
- [Developers](https://aditya-deokar.me/developers): Developer portal, playground, and API key generation
- [Connect](https://aditya-deokar.me/connect): 1-click Claude Desktop & Cursor MCP installation
- [Privacy Policy](https://aditya-deokar.me/privacy): Data handling and zero-tracking commitment

---

## Featured Production Systems

### Full-Stack & AI Systems

- **[a8n](https://aditya-deokar.me/projects/a8n)** — *AI Workflow Automation Platform*: Visual DAG automation platform with durable Inngest step-function execution, 10 custom node types with Handlebars context propagation, and a 22-tool MCP server for AI-agent orchestration. Dual API surface: tRPC for web UI, MCP for AI clients. Tech: Next.js 16, tRPC v11, Inngest, MCP, React Flow.
- **[Artistry Cart](https://aditya-deokar.me/projects/artistry-cart)** — *Service-Oriented E-Commerce SaaS*: Multi-vendor e-commerce platform built as an Nx monorepo with 9 deployable services, API gateway routing, Kafka analytics pipeline with DLQ, Stripe Connect payments, and an isolated AI Vision service using Gemini and CLIP. Tech: Node.js, Express, MongoDB, Kafka, Redis, Kubernetes, Docker, Gemini.
- **[Verto AI](https://aditya-deokar.me/projects/verto-ai)** — *AI-Native Presentation Workspace*: Productivity workspace with an 8-agent LangGraph pipeline, hosted MCP server (11 tools), BYOK multi-provider runtime, and Inngest-powered mobile design generation. Layout selection before content writing ensures structure-aware presentation output. Tech: Next.js 16, LangGraph, Inngest, MCP, Gemini 2.5, BYOK.
- **[NeoPrep AI](https://aditya-deokar.me/projects/neoprep-ai)** — *AI Career Preparation Platform*: Education technology platform with streaming course generation via Gemini, voice-driven mock interviews through Vapi, and Inngest-scheduled weekly industry intelligence on Next.js 15. Tech: Next.js 15, TypeScript, Gemini, Vapi Voice AI, Vercel AI SDK, PostgreSQL.

---

## Core Technologies & Competencies

- **Frontend Architecture**: React, Next.js (App Router, Server Actions), TypeScript, GSAP, Three.js, Framer Motion, Tailwind CSS
- **Backend & API Systems**: Node.js, Express, tRPC, REST (v1), GraphQL, Inngest (Durable Execution)
- **AI / ML & Agent Systems**: LangGraph, Model Context Protocol (MCP), Gemini 2.5, OpenAI GPT-4, Vercel AI SDK, Vapi Voice AI
- **Data & Caching Layer**: PostgreSQL, MongoDB, Prisma, Redis
- **Infrastructure & DevOps**: Docker, Kubernetes, Kafka, Vercel, AWS

---

## Official CLI Tool

Run without installation via `npx`:
\`\`\`bash
# Inspect Aditya's profile
npx adityadeokar profile

# List production projects and stacks
npx adityadeokar projects

# Read deep architecture case study
npx adityadeokar case-study verto-ai

# Output MCP setup for Claude Desktop & Cursor
npx adityadeokar mcp
\`\`\`

---

## Social & Contact

- **Email**: adityadeokarpc@gmail.com
- **GitHub**: https://github.com/aditya-deokar
- **LinkedIn**: https://www.linkedin.com/in/aditya-deokar-4035b5221
- **X (Twitter)**: https://x.com/DeokarAdit31240

---

## Machine Protocol Details

- **REST Rate Limits**: Standard RFC RateLimit headers returned (`RateLimit-Limit`, `RateLimit-Remaining`, `RateLimit-Reset`, `RateLimit-Policy`).
- **REST Versioning**: `/api/v1/` routes with `API-Version: 1.0.0` and RFC 8594 Sunset policy.
- **MCP Streamable HTTP**: `POST https://aditya-deokar.me/api/mcp` speaks JSON-RPC 2.0 (protocolVersion: 2024-11-05).
- **Zero-Auth Key Generation**: `POST https://aditya-deokar.me/api/keys` returns an instant live developer API key.