Zahid hasan@Zahi49

PocketForge AI is an autonomous AI Dark Factory testbed and simulation pipeline designed to build, stress-test, self-repair, and cryptographically verify mission-critical software without human intervention. Governed by the core philosophy "AI that builds software, breaks it, fixes it, and proves it," the system automates the entire software lifecycle inside a deterministic closed loop. The architecture operates through a 6-agent orchestration pipeline using strongly typed handoffs: Planner Agent: Decomposes ingress task briefs into acceptance criteria, failure hypotheses, and edge-case models. Architect Agent: Defines database schemas, state machines, and mathematical conservation invariants. Implementation Agent: Synthesizes backend business logic and integer minor-unit financial accounting. Adversarial Test Agent (Red Team): Executes a 50-vector automated exploit suite targeting race conditions, duplicate retries, and network cutoffs. Repair Agent: Diagnoses failure traces and injects atomic mutex locks and rollback invariants. Verification Agent (Blue Team): Acts as an independent zero-trust auditor, validating ledger consistency and generating sealed SHA-256 evidence proof hashes. Its reference demo engineers an ultra-resilient wallet transfer engine guaranteeing zero fund loss, race-condition immunity, and instant idempotency across thousands of concurrent attacks.

AURA AI Trading Agent is an advanced, autonomous options trading and risk intelligence console designed to empower modern traders. Navigating the financial markets requires split-second decision-making and deep analytical capabilities, which is exactly what AURA delivers. Our platform provides a simulated, risk-free environment where users can test strategies, track real-time portfolio balances, and leverage artificial intelligence to analyze complex market trends. At the core of AURA is a powerful AI engine that evaluates risk-reward ratios and autonomously executes simulated trades based on data-driven insights. What truly sets this project apart is its robust, highly scalable microservices architecture. The system is intentionally decoupled into three distinct layers: a lightning-fast user interface built with Vue.js (hosted on Vercel), a secure core backend powered by Laravel, and a dedicated AI processing engine built with Python and FastAPI. The backend services are containerized using Docker and deployed on Render. By separating the frontend, business logic, and heavy AI computations, AURA ensures high availability, fault tolerance, and seamless scalability. This project successfully bridges the gap between sophisticated financial modeling and user-friendly software design, offering a comprehensive tool for both novice learners and fintech enthusiasts looking to explore the future of algorithmic trading.

Inbox2Order AI is an AI-powered order automation platform designed to help small businesses manage customer orders more efficiently. Many businesses receive orders through WhatsApp, Instagram, SMS, and email, where important information is often buried inside unstructured conversations. Inbox2Order AI transforms these messages into structured order data by extracting details such as customer information, products, quantities, delivery addresses, payment methods, and other order requirements. The platform provides a centralized dashboard for managing orders, monitoring order status, tracking inventory, and viewing business insights. Its AI-based information extraction workflow reduces repetitive manual data entry and helps minimize human errors. Inbox2Order AI is built using React, TypeScript, and Vite with a responsive, modern interface. The project was developed for a hackathon with a strong focus on solving a practical business problem through a functional AI-powered workflow rather than simply presenting a concept. Future development can extend the platform with direct WhatsApp and Instagram integrations, automated notifications, CRM integration, payment systems, and more advanced business automation.

BandFix AI is an autonomous multi-agent debugging system designed to revolutionize how developers fix code errors. Instead of manually searching through logs and stack traces, users simply submit buggy code and error messages into the system. The platform then activates a coordinated AI agent squad consisting of an Orchestrator, Bug Finder, Fix Generator, and Reviewer. Each agent performs a specialized role in the debugging pipeline. The Orchestrator manages task flow, the Bug Finder identifies root causes, the Fix Generator produces corrected code, and the Reviewer ensures security and performance quality. All agents communicate through a real-time event-driven system using a custom message bus, ensuring transparency and structured execution flow. Users can visually track each stage of debugging in real time through a modern interactive dashboard. BandFix AI eliminates manual debugging effort, reduces development time, and introduces a scalable approach to autonomous software engineering. It transforms debugging from a reactive process into an intelligent, automated pipeline. This project demonstrates the future of AI-assisted development where multiple specialized agents collaborate like a real engineering team.

In today's fast-paced financial markets, retail investors often lack access to the rapid, data-driven insights available to institutional players. MarketPulse AI bridges this gap by acting as a highly intelligent, autonomous financial analyst. Our platform seamlessly transforms scattered web signals—such as real-time news sentiment, hiring velocity spikes, and institutional movements—into clear, actionable intelligence. Built with a dynamic Next.js frontend and a robust FastAPI backend, MarketPulse AI offers a comprehensive suite of analytical tools. Users can track their favorite stock tickers, receive predictive pre-earnings scores, and evaluate risk and confidence metrics at a glance through an intuitive dashboard. Furthermore, the platform features an interactive AI chat for deep-dive market analysis and automatically generates detailed, downloadable intelligence reports (PDF/CSV). By combining multiple data dimensions into a single ecosystem, MarketPulse AI empowers users to make smarter, data-backed investment decisions before the broader market even reacts.

VoiceBroker AI is an autonomous voice-powered trading platform built for the AI Agent Olympics Hackathon 2026 at Milan AI Week. The platform enables users to control a simulated trading desk entirely through natural voice commands. Users can speak commands such as “Buy $50 of Apple stock” or “Sell my Tesla position,” and the system automatically processes the request using real-time AI pipelines. The application combines Speechmatics real-time speech-to-text, Google Gemini AI for intent parsing and market reasoning, Kraken market APIs for trading data, and Supabase for portfolio management and historical tracking. VoiceBroker AI demonstrates a complete autonomous agent workflow: Voice Input → Speech Recognition → AI Intent Analysis → Risk & Market Evaluation → Trade Execution → Voice Confirmation The system includes intelligent fallback architecture to ensure reliability during live demos. If Gemini APIs fail or rate limits are reached, the platform automatically switches to rule-based parsing systems powered by real operational data. Key features include: • Real-time voice-controlled trading • AI-powered trading analysis • Autonomous trade execution simulation • Portfolio intelligence dashboard • Market forecasting insights • Voice confirmation system • Real-time AI processing visualization • Enterprise-grade fallback reliability The project demonstrates how multimodal AI agents can transform financial operations by making trading conversational, autonomous, and accessible. Tech Stack: React, TypeScript, Tailwind CSS, Supabase, Speechmatics, Google Gemini 1.5 Flash, Kraken APIs, Recharts, and Framer Motion.

Nexova AI is a full-stack enterprise intelligence platform designed to help modern businesses monitor operations, analyze revenue, manage inventory, forecast trends, and interact with company data through a powerful AI assistant. The platform combines real-time business analytics with AI-driven insights inside a modern enterprise dashboard experience. Users can track revenue, orders, contracts, inventory levels, and forecasting data through interactive visualizations and intelligent automation workflows. One of the core innovations of Nexova AI is the integrated AI Agent system powered by Google Gemini 1.5 Flash. The AI assistant can analyze operational data, answer questions about revenue trends, detect inventory risks, summarize contracts, generate reports, and provide forecasting insights using real business context from Supabase. The application was designed with resilience and demo reliability in mind. If external AI services fail or API limits are reached, the system automatically falls back to intelligent rule-based responses using real Supabase business data. This ensures the application never crashes during demonstrations or production scenarios. Key features include: - Enterprise analytics dashboard - AI-powered operational assistant - Revenue forecasting engine - Inventory intelligence and stock alerts - Contract analysis with AI summaries - Real-time Supabase integration - PDF/export/share workflows - Responsive SaaS UI with premium animations The platform uses React, TypeScript, Vite, Tailwind CSS, Supabase, Gemini AI, Recharts, Framer Motion, and Lucide React to deliver a production-style enterprise experience. Nexova AI demonstrates how AI agents can transform enterprise operations by making business intelligence more accessible, conversational, and actionable.

Sentinel.AI is a futuristic AI-powered cybersecurity intelligence platform designed to simulate autonomous vulnerability analysis and secure document scanning workflows. The platform demonstrates how modern AI agents can assist cybersecurity teams by automating attack-surface analysis, vulnerability reasoning, and threat reporting through intelligent workflow orchestration. Sentinel.AI includes two primary modules: 1. AI Vulnerability Scanner Users can scan web applications, APIs, network targets, and system environments through an advanced AI-inspired scanning interface. The platform simulates intelligent reasoning pipelines that enumerate attack vectors, cross-reference CVE/CWE databases, apply OWASP heuristics, calculate risk severity, and generate triaged security reports. 2. SecureDoc AI An AI-powered PDF and document security analyzer capable of detecting suspicious links, hidden scripts, malware indicators, credential leaks, phishing patterns, and unsafe attachments. The workflow demonstrates autonomous threat classification and ML-inspired reasoning. The application uses futuristic live-trace visualizations, AI workflow simulation, neural threat analysis concepts, and cyberpunk-inspired UI/UX design to create a realistic next-generation security operations experience. Sentinel.AI was built as a frontend-first prototype demonstrating: - Agentic AI workflows - AI-assisted cybersecurity automation - Autonomous threat reasoning - ML-inspired classification systems - AMD GPU-ready architecture concepts - Scalable cybersecurity intelligence pipelines Tech Stack: - React - TypeScript - Lovable AI - AMD Developer Cloud concepts - ROCm-inspired workflow architecture - AI Security Intelligence UI The project highlights how intelligent AI systems can improve cybersecurity workflows through automated reasoning, real-time threat analysis, and scalable compute-assisted intelligence pipelines.

Prompt Pay is a real-time per-API monetization system built for the Agentic Economy using Arc’s nanopayment infrastructure. It allows developers to charge as little as $0.001 per AI API call, with instant onchain settlement in USDC. Traditional API monetization models fail due to high gas fees, making microtransactions unprofitable. Prompt Pay solves this by leveraging Arc’s feeless or ultra-low-cost transaction system, enabling true pay-per-use economics for AI services and machine-to-machine interactions. The platform includes a test wallet system, per-call pricing enforcement, real-time transaction logs, and a bulk testing environment to simulate multiple API calls. Each request is only processed after successful payment confirmation, ensuring a secure payment-first access model. This project demonstrates how AI APIs, agents, and digital services can operate under a scalable microtransaction model. It highlights the future of decentralized monetization, where every action has value and can be settled instantly without traditional financial friction.

🌎 Online Hackathon | 💻 Online Build: Sept 26 - Oct 5, 2026 A virtual, global hackathon. Build from anywhere in the world. Dark Factory is the official hackathon of WeAreDevelopers World Congress North America, presented by BAND, but you don't need to be in San Jose to take part. For one week, you build a software factory in BAND Desktop: a band of coding agents that plans work, implements it and checks its own results. 🎟️ Bonus for sign-ups: you are eligible for a free ticket to WeAreDevelopers World Congress North America in San Jose (Sep 23–25). If you do go, stop by the BAND booth: a surprise is waiting for you.

🌎 Online Hackathon | 💻 Month-long challenge | Dates: Sep 1–30, 2026 | 🏆 $10,000 Prize Pool ($5k cash + $5k in AAI credits) A month-long online challenge run by lablab.ai together with AssemblyAI, the Voice AI infrastructure company for builders. Every participant builds on AssemblyAI. You can join at any point — registration stays open for the whole build window, so you can start on day one or pick the project up halfway through and still submit. 🌍 Fully online hackathon. Join and build from anywhere in the world.

🌎 Online Hackathon | 💻 Dates: 28 August–4 September 2026 | 🏆 $6,000 Prize Pool Build AI trading agents on Alpaca — autonomous agents and trading apps using Alpaca's Trading API, MCP server and CLI. 🌍 Fully online hackathon. Join and build from anywhere in the world.