Obaid Ullah Safi@obaid_ullah87

Autonomous trading agents have historically chosen between two failure modes: execute freely with no guardrails (high-performance, high-risk), or sit behind policy engines that never actually trade (safe but inert). This gap is quantifiable — H1 2025 crypto theft alone reached $3.01 billion. PolicyMint closes both sides of the problem: it trades actively while proving every risk decision on-chain through the ERC-8004 Validation Registry, making institutional-grade safe execution achievable for the first time. How the Policy Engine Works Before every trade execution, PolicyMint runs a mandatory pre-flight evaluation against three core policy rules: Spend Cap — Per-transaction USD limit capped at ≤ $450 to stay below the RiskRouter's $500 hard ceiling Venue Allowlist — Only whitelisted trading venues (e.g., kraken-spot, uniswap-v3) are permitted to receive orders Daily Loss Budget — Cumulative drawdown tracked in real time; agent halts when the daily loss threshold (≤ 4% of capital) is breached Every trade intent is structured as an EIP-712 typed-data payload, cryptographically signed with the agent wallet, and bound to Ethereum Sepolia (chainId 11155111) via EIP-155. If the policy engine returns allow, the signed intent becomes the execution proof submitted to the ERC-8004 RiskRouter. If the engine returns block, the Kraken CLI call never fires and a human-readable block reason is logged.

SmartFit is an AI-powered recruitment platform that automates the entire candidate evaluation process using two advanced Opus workflows integrated with a clean React Vite frontend. It focuses on solving the biggest challenges in hiring messy resume formats, inconsistent evaluations, bias, and slow decision-making by converting unstructured documents into structured, explainable, and actionable insights. The system begins by extracting detailed information from both candidate CVs and job descriptions, no matter the format: PDFs, DOCX files, images, screenshots, or plain text. Using a combination of OCR, semantic parsing, and AI agents, SmartFit generates structured profiles that include skills, experience, education, projects, responsibilities, and more. This enables reliable comparison between candidates and job requirements. Once the data is extracted, SmartFit applies a hybrid scoring engine that mixes rule-based validations with deep AI reasoning. It scores each candidate across multiple dimensions such as skills, experience relevance, education alignment, language proficiency, and cultural fit. The output is a transparent, auditable evaluation report that explains exactly how each score was calculated and which rules were triggered. The second workflow performs semantic job matching, identifying which skills the candidate already possesses, which are missing, and which are extra strengths. It also generates personalized recommendations that candidates or HR teams can use to improve CVs, prepare for interviews, or shortlist talent more confidently. The frontend ties everything together by offering an easy interface where users can upload documents, run workflows, and instantly view structured outputs, scores, fits, and recommendations. SmartFit ultimately reduces HR workload, improves accuracy, and brings transparency and fairness to the hiring pipeline. It is built for scalability and designed to handle real-world hiring scenarios.

GigStream is an innovative AI-powered real-time USDC payment streaming platform designed to revolutionize the gig economy, built on Circle's Arc blockchain. In a world where gig workers face frustrating payment delays—often waiting days or weeks for earnings—GigStream eliminates this pain by automating instant micro-payments as tasks are completed. Leveraging AI for intelligent task verification, risk scoring, and predictive financial optimization, the platform ensures sub-second settlements directly to workers' wallets, fostering financial stability and inclusion. At its core, GigStream offers real-time payment streaming, where workers get paid per task or per minute, not monthly. AI-driven predictive micro-loans allow safe advances on future earnings, auto-repaid from upcoming tasks, with risk-adjusted fees. An on-chain reputation system provides transparent, portable credit scores, enabling workers to build trust across platforms. Additional features include automated tax and expense tracking for compliance, multi-wallet management for flexible fund routing, and dynamic job routing for optimized opportunities. Technically, the monorepo architecture spans smart contracts (Solidity on Arc), a backend (Node.js, Cloudflare Workers, PostgreSQL), and a frontend (Next.js, React). Circle's Developer-Controlled Wallets handle secure, gasless USDC transfers, while AI models run on Cloudflare Workers for edge computing.

Searching and applying for jobs is often a tiring and repetitive process. Candidates spend countless hours scanning job boards, adjusting resumes, and drafting cover letters, which can be overwhelming and inefficient. Many great opportunities are missed simply because people cannot keep up with the workload. Our project tackles this by building an AI Job Application Agent powered by a modular multi-agent system. Each agent has a dedicated role: parsing resumes, searching job listings through APIs, ranking positions based on profile fit, and generating personalized cover letters with an LLM. A lightweight user interface ensures that candidates remain in control, they can review, approve, and decide whether to submit applications. The design emphasizes transparency and accountability, with every step logged. By combining automation with a user-in-the-loop approach, we make job applications both smarter and more efficient. Beyond the hackathon, the system can evolve into a premium service with integrations for LinkedIn, ATS platforms, multilingual support, and adaptive feedback loops that learn from successful applications. In short, our solution makes job hunting less stressful and more strategic, helping candidates focus on opportunities that truly matter.

ShopGPT, powered by GPT-5, is a cutting-edge tool that transforms online shopping by delivering personalized, high-quality product recommendations. Its ability to analyze user preferences, filter out low-quality products, and prioritize those with strong reviews makes it a valuable asset for both consumers and businesses. By saving time, reducing costs, and ensuring premium quality, ShopGPT enhances the shopping experience while helping retailers optimize for AI-driven discovery. To maximize its benefits, users and businesses should focus on clear queries, structured data, and maintaining positive reviews.

Athena solves the collaboration challenge faced by development teams, especially beginners intimidated by Git workflows. Our platform uses AI agents to create a seamless multi-developer experience: the Overview Agent helps users refine project requirements, the Code Generation system builds full-stack applications, the AI Orchestration Agent manages change requests to prevent merge conflicts, and the Code Review Agent onboards new developers instantly. Multiple team members can work on the same project using a common project ID, eliminating the need for Git knowledge while maintaining professional development standards. Athena democratizes collaborative software development, making it accessible to teams of all skill levels.

Gaia is an innovative web application designed to provide safety and support to women who find themselves alone in potentially risky situations. The app can chat using an AI to give the user the illusion of being in company, offering both emotional reassurance and a sense of security. In addition, Gaia can help users call emergency services instantly if they are in danger. The app also features a map that identifies the most dangerous areas based on real-time emergency call data, enabling users to avoid risky locations. This data-driven approach will also assist law enforcement in intelligently focusing their efforts on areas with higher safety concerns. We have worked just for London and we have got the data from their official website.

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.

⏱️ 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.

📅 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.