Pavan Kumar Ponnuri — AI-Native Software Developer · Gen AI Engineer · Freelancer · Visakhapatnam, India 🇮🇳
I design and build backend-heavy, AI-native software — from production-grade platforms to autonomous multi-agent systems. My focus is on systems that are simple to use, hard to break, and built to scale.
- 🔨 Currently building a Social Media & HR Agent, Email Agent, and PC Agent using the Hermes agent framework
- 🎯 Core focus areas: Backend APIs · AI Agents · Generative AI Products
- 💼 Open to freelance projects, full-time roles, and GenAI collaborations
| Domain | Technologies |
|---|---|
| Languages | |
| Backend | |
| Full Stack | |
| Databases | |
| AI · ML · GenAI | |
| Open Source AI | |
| DevOps & Tools |
What makes it special: 🛡️ Anti-cheat system — fullscreen enforcement, tab-switch auto-disqualification, paste detection, DevTools blocking, and remote disqualify by admin in real time 📝 Two assessment types — Output Prediction (code snippet → predict output) and MCQ, both auto-graded server-side 🔁 Typing replay — keystroke timeline playback with Monaco Editor viewer and paste markers for malpractice review 📚 Question Library — reusable bank; import into any round in one click with auto-distributed test case points 📊 CSV export — all results or cutoff-filtered finalists; set passing threshold per round 🔐 Admin auth — Supabase Auth with Google OAuth and Magic Link; no candidate login required
|
What makes it special: 🔐 Local JWT Auth — customers sign up with email + password (scrypt-hashed); OTP email verification on signup, login, and checkout; admin secured via HMAC-signed JWT 🛍️ Cart, buy-now flow, OTP-verified checkout, and UPI payment integration with branded PDF quotations 📦 Item-level delivery tracking with real-time order status updates 💬 In-order messaging between customers and admin 🛠️ Full admin dashboard — products, orders, payments, customers, analytics, and audit logs
📐 System Architecture |
What makes it special: 🎬 Three input modes — video upload, audio upload, or any public YouTube URL (auto-downloaded via yt-dlp)
📝 Timestamped transcription via Sarvam.ai
|
What makes it special: 🌊 Live SSE streaming — AI output rendered token-by-token in real time with a masked preview window 🤖 Groq Llama 3.3 70B generates tailored day-by-day itineraries with restaurant picks, activities, and practical info 🗺️ Interactive Leaflet.js maps — colour-coded per-day pins with Google Maps deep-links; no API key required 💱 30 currencies via Frankfurter API with live FX rate annotations inline in the itinerary text 🎒 AI packing list — emoji-categorised checklist generated per destination, travel style, and accessibility needs ♿ Accessibility-first filters — wheelchair, elevator, limited-walking, visual and hearing impairment options
|
┌───────────────────────────────────────────────────────────────────┐
│ │
│ 🧩 Design First — Architecture before code │
│ 🔁 Iterate Fast — Ship MVPs, refine in production │
│ 🔌 API-Driven — Every feature starts as an endpoint │
│ 🤖 AI-Augmented — LLMs as first-class system components │
│ 📏 Measure Everything — Logs, metrics, observability by default│
│ 📦 Own the Stack — Backend to frontend to deployment │
│ │
└───────────────────────────────────────────────────────────────────┘
Generated daily by Platane/snk — orange trail on the profile's purple palette
🏁 START ═══════════════════════════════════════════════ 🏆 DESTINATION
║ ║
║ 🚗💨 ───── Backend ───── Full-Stack ───── GenAI ─────► ║
║ Mastery Products Agents ║
║ ║
╠══════════════════════════════════════════════════════════════╣
║ ║
║ 1. System Design 2. Scalable Platforms ║
║ Distributed systems Backend-heavy web products ║
║ ║
║ 3. AI Agents 4. Freelance Practice ║
║ LLM-powered tools Shipping quality for clients ║
║ ║
╚══════════════════════════════════════════════════════════════╝
- 💼 Freelance and contract work — backend APIs, full-stack platforms, AI integrations
- 🏢 Full-time roles — backend, full-stack, and AI-native engineering
- 🤖 GenAI collaborations — agent frameworks and LLM-powered products
- 💬 Technical discussions — system design, architecture trade-offs, and scaling
- 🌍 Open source — backend, AI, and developer-tooling projects


