Tanya Garg@Tanya5315

Overview: TriadInspect AI is an AMD-powered industrial inspection copilot designed to help factory teams identify potential equipment issues by combining machine images, sensor readings, and technical manuals in one workflow. Problem Statement: Industrial maintenance teams often analyze visual inspections, temperature and vibration readings, and equipment manuals separately. This can make it difficult to detect inconsistencies, verify potential faults, and prepare maintenance reports efficiently. Proposed Solution: TriadInspect AI cmbines these evidence sources to support more informed maintenance decisions. An AI model running on AMD-powered infrastructure analyzes the available information, while a rule-based verification engine checks sensor values against documented thresholds and highlights missing or conflicting evidence. Key Features: 1. Machine inspection using available visual evidence. 2. Sensor monitoring for readings such as temperature and vibration. 3. Technical manual and maintenance checklist analysis. 4. Rule-based validation of numerical thresholds. 5. Inspection statuses: Monitor, Inspect, and Escalate. 6. Maintenance work-order drafts and downloadable reports. 7. Transparent findings that support human review. Technology Stack: Python, Streamlit, AMD-powered AI inference, Pandas, Plotly, and document-processing tools. Expected Impact: TriadInspect AI aims to simplify industrial inspection workflows, improve the traceability of findings, and reduce the effort required to prepare maintenance reports. It combines AI analysis with deterministic checks to make recommendations more transparent. The prototype will be tested using normal, abnormal, incomplete, and conflicting evidence scenarios. It is designed to assist maintenance teams, not replace expert judgment or certified safety inspections.

TableGuard is an autonomous AI Dark Factory built in BAND Desktop to create a reliable restaurant reservation service through coordinated coding agents. Our factory uses three specialized seats: an Architect, a Builder, and an Independent Verifier. The Architect analyzes requirements, identifies edge cases, and creates an implementation plan. The Builder converts that plan into working software and produces evidence. The Verifier independently tests and attacks the implementation rather than simply trusting the Builder. The core reliability challenge is preventing double-booking when multiple requests compete for the same table. TableGuard also focuses on retries, overlapping reservations, invalid inputs, and time-zone edge cases. When verification discovers a defect, the Builder receives the finding, reproduces the issue, fixes it, and the Verifier tests the correction again. Our workflow is: PLAN → BUILD → ATTACK → REPAIR → VERIFY The goal is not simply to use AI to generate code, but to demonstrate a reusable software factory where coding agents collaborate autonomously, challenge each other's work, recover from failures, and produce evidence before accepting a result.

NPCVerse AI is an AI-powered platform that transforms traditional game NPCs into intelligent, memory-driven digital characters capable of learning, adapting, and evolving through every player interaction. Instead of relying on repetitive dialogue trees and scripted behaviors, NPCVerse creates lifelike NPCs with persistent memory, dynamic personalities, emotional intelligence, and context-aware decision-making. Every conversation, quest, and player choice shapes how an NPC remembers, reacts, and responds, making each gameplay experience unique and immersive. The platform is powered by a multi-agent architecture where specialized AI agents manage different aspects of character intelligence. A Personality Agent defines traits, goals, and communication style, while a Memory Agent stores conversations and important events. An Emotion Agent updates feelings such as trust, fear, or happiness based on player actions. A Dialogue Agent generates natural, context-aware conversations, and a Behavior Agent decides how the NPC should respond, assign quests, trade, or interact with players. A World Knowledge Agent ensures every character remains consistent with the game's lore and environment. Built using AMD Developer Cloud, Fireworks AI, open-source language models, and Docker, NPCVerse delivers scalable, real-time AI experiences that integrate easily with modern game engines. Our vision is to replace static scripted NPCs with believable digital personalities that remember the past, react to the present, and evolve for the future, enabling developers to create richer stories, stronger player engagement, and truly living virtual worlds.

InvestiBand is a Band-powered adversarial startup due diligence platform that helps investors evaluate startups through collaborative AI agents instead of relying on a single AI model. A user simply enters a startup name, and multiple specialized agents join a shared Band room to investigate the company together. The Research Agent gathers information about founders, funding rounds, company background, and public data sources. The Market Agent analyzes competitors, industry trends, market size, and growth opportunities. The Finance Agent evaluates the business model, revenue streams, unit economics, and profitability potential. To encourage balanced decision-making, InvestiBand introduces two adversarial agents: Bull and Bear. The Bull Agent argues why the startup is a strong investment opportunity, while the Bear Agent highlights weaknesses, risks, and reasons not to invest. Their debate happens live inside the Band room, making the reasoning process transparent and easy to understand. A Risk & Verification Agent continuously checks claims made by other agents, flags unsupported statements, and can veto conclusions that lack evidence. This creates a more reliable and auditable workflow. Finally, a Human Approval Agent reviews the discussion and produces an investment memo with scores for market, finance, risk, and team quality, along with a final recommendation: Invest, Wait, or Reject. Band is the core collaboration layer of InvestiBand, enabling agents to exchange context, debate ideas, verify facts, and coordinate decisions in real time.

Six Matches. One leaderboard. Build on AMD, earn points and compete for prizes.

🌎 Hybrid Hackathon | 💻 Online Build: 12-17 October 2026 | On-site phase: 17-18 October 2026 To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account. The AMD Developer Hackathon: ACT III is a hands-on event for developers, founders, engineers and builders who want to push what’s possible with AI on real infrastructure. 📍 If approved for the on-site experience, you'll also have the opportunity to showcase your project live. Travel and accommodation will not be covered.

🌎 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 | ⏳ Build: 48 hours | 📅 September 25–27, 2026 The official IBM Bob 2.0 hackathon. Build what's next in AI-assisted development, alongside the developers and experts shaping it. ⭐ Get help from experts throughout the event 🤝 Join a team or build solo 💰 Prize pool: $12,000 🧑🏻‍💻 Register before the Kick-Off Stream to start building with Bob