Muhammad Talha@muhammadtalha

The migration of high-performance computing workloads from NVIDIA's CUDA to AMD's ROCm ecosystem is a critical bottleneck for developers adopting new hardware. Ghost-Coder eliminates this friction by serving as an autonomous, self-healing translation agent. Powered by the Qwen2.5-Coder-32B model running on an AMD Instinct™ MI300X (192GB) instance, Ghost-Coder doesn't just do static text replacement—it deeply analyzes device-side kernels and host-side memory API calls to generate native, highly optimized HIP code. Architecture & Workflow: The system is built with a decoupled frontend/backend architecture to maximize GPU resource allocation. The UI is a reactive Gradio web application hosted via Hugging Face Spaces, designed with an agentic interface that visualizes the translation and verification steps in real-time. This frontend communicates via a secure tunneling protocol to a custom FastAPI bridge running directly on the AMD Developer Cloud droplet. Inside the droplet, the AI engine operates under strict hardware guardrails. To achieve stable, high-throughput inference on the MI300X architecture, we engineered a specialized execution environment. By overriding the ISA to gfx942, strictly serializing kernel execution, and disabling experimental Triton memory paths, we stabilized the Qwen model to prevent memory aperture violations during continuous generative decoding. Self-Healing & Translation: When a user submits a CUDA kernel (e.g., a tiled matrix multiplication), Ghost-Coder analyzes the computational logic. It dynamically maps NVIDIA-specific host calls like cudaMalloc and cudaMemcpy to their hipMalloc and hipMemcpy equivalents, while restructuring execution macros (like transforming <<<blocks, threads>>> into hipLaunchKernelGGL). The result is drop-in ready C++ code tailored for the AMD stack. Ghost-Coder accelerates the transition to open-source GPU computing by turning days of manual kernel porting into a seamless, automated seconds-long process.

Vigil is an intelligent risk assessment platform powered by 8 specialized AI agents that work in sequence to analyze financial and operational risks. Users provide company profile data, then Vigil's orchestrator routes the message through agents (Signal Harvester, Narrative Intel, Macro Watchdog, Competitive Intel, Risk Synthesizer, Strategy Commander, Market Oracle) based on intent. The system ingests live market data (VIX, S&P500, sector ETFs, news headlines) and synthesizes it with company context to produce: (1) risk score 0-100 with tier color (GREEN/YELLOW/ORANGE/RED), (2) top 3 risks ranked by probability, (3) top 3 actions with deadlines, and (4) 30-day strategic playbook. Built on FastAPI backend with session persistence, Claude models via AIML API on complete.dev environment.

FLOE revolutionizes blockchain payments by combining Arc Testnet, Circle USDC/SDK, autonomous AI agents, and voice control into one seamless platform for real-world asset automation. We built Agent-to-Agent (A2A) autonomous payments where Cloudflare LLaMA 3 AI agents automatically evaluate payment requests, analyze contract terms, check wallet balances, and execute transactions via Circle SDK—zero human intervention required. Full CCTP integration enables seamless cross-chain USDC transfers between Arc Testnet, Base Sepolia, and Polygon. Multi-wallet management with Circle's Web3 Services SDK handles smart contract execution and real-time transaction polling. Our Voice AI Assistant uses Cloudflare Whisper for speech-to-text, LLaMA 3 for conversational AI, and ElevenLabs for text-to-speech. Users control payments with natural language: "Show my wallets", "Create contract", "Send payment"—the AI navigates and executes instantly. Real use case: Landlord creates $1000 monthly rent request → Enables A2A mode → Sends to tenant → Tenant's AI agent evaluates on due date → Checks wallet balance → Verifies contract terms → Auto-executes payment via Circle SDK → Landlord receives USDC on Arc Testnet. Completely autonomous. Technical highlights: Request Center for payees to receive payments, A2A manual mode (agent evaluates, human executes), A2A auto mode (agent evaluates + executes), recurring payment scheduling, cross-chain CCTP transfers, voice-driven page navigation, production-ready smart contract execution on Arc Testnet. Built with Arc Testnet, Circle USDC & SDK v1.1.1, CCTP, Cloudflare Workers AI, ElevenLabs, Node.js, React, TypeScript, PostgreSQL. Fully functional and ready for real-world RWA payment automation.

QuizWhiz is a cutting-edge learning platform designed to empower students by transforming their study material into interactive learning formats — instantly. Students simply upload a PDF, DOC, or paste a YouTube educational lecture, and our AI (powered by Novita LLMs) analyzes the content and generates: Multiple-choice quizzes Interactive flashcards (with TTS playback) Shareable quiz links for peer testing Smart feedback with correct answers and explanations What makes QuizWhiz truly stand out: No login required: Students can generate quizzes instantly. Voice-enabled flashcards: Turn studying into an audio experience. Save and share: Easily save question sets or challenge friends. Mockup features preview: Group quiz rooms, chat discussion, and personalized study plans. Built using ReactJS, NodeJS, MongoDB, and deployed via Render, this app is fully scalable and fast. We developed this project using TRAE AI (AI coding IDE), which dramatically boosted our speed and consistency. LLM integration via Novita AI APIs played a critical role in converting raw text to meaningful learning content. QuizWhiz is more than a tool — it’s a glimpse into the future of education. Designed with accessibility, speed, and utility in mind, it's perfect for students, teachers, and educational institutions looking to upgrade their learning workflows using AI. With future plans to include group study rooms, spaced repetition memory coaching, leaderboards, and user login dashboards — QuizWhiz is just getting started. Let’s make learning smarter, faster, and more fun — for everyone.

CodeSentinel is an AI-driven code security and review platform that integrates directly with GitHub and Bitbucket to analyze pull requests in real time. Powered by Claude Sonnet, it deeply inspects code diffs to detect logic bugs, security flaws (like SQL injection, insecure auth), and style violations. Unlike traditional scanners, CodeSentinel reasons about code contextually, explains vulnerabilities, suggests secure fixes, and even allows developers to chat directly with the AI to ask follow-up questions or request refactors. I built the entire platform solo in under 4 days — full-stack (React, FastAPI, PostgreSQL), with secure OAuth integration, PR diff handling, AI prompt engineering, and a live chat interface. Developers can review and fix PRs with LLM assistance, right from their dashboard. CodeSentinel elevates code quality, automates security, and saves teams countless hours — all through intelligent, explainable AI. It’s not just a static scanner; it’s a full-fledged AI copilot for your codebase.

📅 May 11–19, 2026 • May 11 – 19: Online Build Phase. • May 18 (Hybrid Build Day – Online & Onsite) • May 19 (Demos & Awards – Online & Onsite) 🌍 Hybrid event. Join online from anywhere or build onsite at the venue. 🏆 Winners will be announced live on stage at the AI & Big Data Expo North America on May 19 📍 On-site Venue (May 18–19): San Jose McEnery Convention Center, CA, USA 🤝 Join solo or with a team. 🎟 FREE AI & Big Data Expo North America ticket included for participants that are able to join the on-site phase. 🏆Compete for a $10,000 prize pool 🧑‍💻 Apply now to design, prototype, and showcase enterprise-ready innovation.

⏱️ Build the next generation of Autonomous Agents in the heart of Europe’s AI Revolution. 📅 May 13 – 20, 2026 May 13 – 19: Online Build Phase. May 19: An exclusive on-site build day at Milan AI Week for selected participants. May 20: Awards Ceremony. 📍On-site Venue (May 19–20): Fiera Milano (Rho), Milan, Italy 🤝 Join solo or with a team. 🎟 FREE Milan AI Week conference ticket included for participants. 💰 $32,000+ prize pool 🧑‍💻 Apply now to turn your autonomous agent into a market-ready enterprise solution.

Build solutions that improve how software is built. Work with an AI that understands your codebase, helping you reduce repetitive work and build with real context. ⏳ Build your project in 48 hours. 🚀 Get hands-on with IBM Bob and explore how AI fits into real development workflows before it becomes standard. 💡 Build tools and workflows that developers would actually use. ⭐ Learn directly from experts throughout the event. 🤝 Join a team or build solo. 💰 Prize pool: $10,000 🧑🏻‍💻 Register before the Kick-Off Stream and start building with Bob.

Grab your ideas, spin up compute, and start building what’s next! Start with $100 in AMD Developer Cloud credits and begin building right away. Get access to hands-on tutorials, workshops, and curated learning resources to help you build faster and go further. To unlock credits and join the hackathon, sign up for the AMD AI Developer Program. 🤝 Go solo or team up — ship something real. 💰 $21,500+ total prize pool & 1 × AMD Radeon™ AI PRO R9700 GPU up for grabs. 📍 On-site Venue (May 9-10): MindsDB SF AI Collective 3154 17th St, San Francisco, California, USA 📍On-site participation is by invitation only. Travel and accommodation expenses will not be covered. 🧑‍💻 Developers, builders, and innovators - bring your boldest ideas and let’s turn them into real AI systems on AMD.

⏱️ Design, build, and deploy trustless AI trading agents. 📅 March 30 – April 12, 2026 🌐 Build AI trading agents that execute strategies using ERC-8004 or Kraken CLI. 🤝 Build solo or in teams to create autonomous trading, risk, yield, or protection agents. 💰 $55,000 prize pool, allocated and shared via Surge and Kraken. Your team should have registered project at early.surge.xyz in order to be eligible for the prize. Get access in your team profile or our discord 🔑 Access to early.surge.xyz If prompted for login credentials, use: • Username: admin • Password: JBRv2xWG7AzwVrLz88

Teams work together with multiple AI agents in shared workspaces to design, build, and launch complete AI products — from initial idea to market-ready app. 📅 February 23– March 1, 2026 ⏱️ 6 days to build and ship an end-to-end agentic AI product 🤝 Work together in shared AI workspaces where your team and AI agents collaborate in real time 🔧 Build with multiple AI agents that handle development, design, testing, and documentation together 🎯 Create everything in one place — from ideation to product to marketing materials, all by prompting and observing 🚀 Focus on real-world products that teams could actually ship and use 🏆 $3,500 in prizes + premium 3-month Complete.dev access for top teams 🧑‍💻 Join your team workspace and start building together