Duy Khang Luong@sunet

ContractForge Auditor is an end-to-end AI governance platform built for enterprises that need to audit contracts quickly and accurately. Users upload a contract (PDF or text) alongside a policy CSV, and the system runs a multi-agent LangGraph pipeline powered by Google Gemini to extract clauses, detect policy violations, score risk across five categories (legal, financial, operational, compliance, data privacy), simulate five risk scenarios (force majeure, penalty delay, data breach, termination, payment default), and generate actionable recommendations. The platform supports bilingual contracts in English and Vietnamese, automatically detecting the language and keeping all AI-generated summaries in the contract's language. A full audit trail with SHA-256 hashes, latency metrics, and PII redaction ensures transparency and compliance. Results are delivered through an interactive React dashboard with a risk score gauge, heatmap, clause explorer, simulation panel, and a downloadable PDF report. Built with FastAPI + LangGraph + Google Gemini on the backend and React 18 + Vite + Tailwind + shadcn/ui on the frontend. Deployed on Render (backend) and Vercel (frontend).

CareFlow Orchestrator is a multi-agent AI platform designed to eliminate the coordination bottleneck in clinical decision-making. In modern healthcare, clinicians must synthesize insights from multiple specialties — radiology, oncology, cardiology, and pharmacy — often working in silos with no unified view. CareFlow solves this by orchestrating four specialized AI agents in parallel, each analyzing the patient case from their domain expertise, then reconciling their findings into a single structured care plan. The workflow begins when a clinician submits a patient case as free-text notes, a medical image (X-ray, CT scan, MRI), or dictated audio via the microphone. The Orchestrator Agent, powered by Gemini AI, decomposes the case and identifies which specialties are relevant. It then dispatches the appropriate Specialty Agents simultaneously — each generating findings, action items, and recommendations specific to their domain. Finally, the Coordinator Agent reconciles all outputs into a unified Care Plan containing a chronological timeline, prioritized recommendations, critical alerts (drug interactions, urgent findings), and detailed findings grouped by specialty. The three-panel React dashboard presents everything in real time: clinicians can watch each AI agent work via the live Agent Chat stream, review the full care plan in the center panel, and export results as a structured PDF or EMR-compatible text file for integration into existing hospital workflows. Built with FastAPI, SQLAlchemy, React, Vite, TypeScript, and Tailwind CSS. Deployed on Render (backend) and Vercel (frontend). Speech input supports both Speechmatics and the browser's native Web Speech API as fallback.

PHI Compliance Scanner is a repo-level PHI (Protected Health Information) scanner built for the IBM Bob Hackathon. It targets a realistic but fully synthetic EHR demo repo with three Python microservices (patient-api, billing-service, audit-logger) and deliberately injected HIPAA-style violations such as logging SSNs, propagating patient identifiers between services, and storing unmasked fields. Instead of hand-writing all the tooling, I used IBM Bob IDE as the primary development partner. Bob read the PRD and .bob/rules.md file, which defines the 18 HIPAA identifiers and scanning rules, then helped plan and implement the scanner/ modules, the PHI flow mapping, the markdown/JSON audit report format, and pytest test stubs. All major steps – planning the synthetic repo, building the scanner engine, generating tests, and doing the final review – are captured as exported Bob task sessions under bob_sessions/. The result is a single-command workflow: python scanner/phi_scanner.py --repo ./ehr-demo --output ./reports/audit-report.md The scanner walks the entire repo, detects and classifies hundreds of PHI violations by service, file and identifier type, builds a PHI flow graph across services, and generates an audit-ready markdown report plus a structured JSON file. Today the focus is fast, accurate detection and reporting; auto-patching, watsonx.ai-based classification, CI/CD integration and dashboards are explicitly designed as roadmap steps.

Access to quality dermatology care is limited in many regions, leaving patients without timely diagnosis for skin conditions and wounds. MediVision addresses this gap by providing an AI-powered assistant that analyzes skin and wound images combined with patient symptom descriptions to deliver structured, actionable insights. Built for the AMD Developer Hackathon 2026 (Track 3: Vision & Multimodal AI), MediVision uses the Qwen2.5-VL-7B vision-language model deployed via vLLM on AMD Instinct MI300X GPUs through the AMD Developer Cloud. The model performs multimodal analysis, processing both images and text to identify conditions such as abrasions, dermatitis, eczema, ringworm, psoriasis, burns, and cellulitis. Key Features: - Multimodal Analysis: Combines skin/wound images with patient symptom descriptions for comprehensive assessment. - Multilingual Support: Outputs in 6 languages — English, Tiếng Việt, 中文, Español, Français, and 日本語 — making it accessible globally. - Structured Output: Provides diagnosis, severity level (Low/Medium/High/Urgent), treatment recommendations, and confidence scores. - AMD Hardware Powered: Leverages AMD MI300X + ROCm for fast, efficient inference. - Gradio Frontend: Hosted on Hugging Face Spaces, providing an easy-to-use interface with minimal dependencies. Tech Stack: - Vision Model: Qwen/Qwen2.5-VL-7B-Instruct - Inference: vLLM (OpenAI API compatible) - Hardware: AMD Instinct MI300X + ROCm - Host: AMD Developer Cloud - Frontend: Gradio 5.29 on HF Spaces Live Demo: https://huggingface.co/spaces/lablab-ai-amd-developer-hackathon/medivision-ai-agent Source Code: https://github.com/HiImSunny/medivision-ai-agent

Learn how modern AI agents interact with the live web through hands-on workshops, real infrastructure, and a community builder weekend in San Francisco. 📅 May 25–31, 2026 • May 25 – 30: Online Build Phase • May 30: Hybrid Build Day (Online & Onsite) • May 31: Demos & Awards (Online & Onsite) 🌍 Hybrid event. Join online from anywhere or build onsite in San Francisco. 🚀 The onsite experience will take place during the opening of Bright Data’s new SF space, The Web Data Loft. 📍 On-site Venue (May 30–31): The Web Data Loft 625 2nd St, San Francisco, CA 🏆 $5,000 cash prize + Bright Data AI Startup Program opportunities 🍕 Free food, drinks, networking, workshops, mentorship, and live demos 🧑‍💻 Apply now to build AI agents powered by real-time web access.

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.