InspectorAI Pro - Generative UI Global Hackathon: Agentic Interfaces
AI Tinkerers - Hyderabad
Hackathon Showcase

InspectorAI Pro

Team led by a Woxsen University researcher and Guinness World Record achiever skilled in Python, GenAI, CNNs, and deploying scalable ML pipelines on Google Cloud.

1 member

The Agentic Engineering Tutor is designed to bridge the gap between complex industrial data and actionable human knowledge by replacing traditional, static dashboards with a generative interface.

Primary Users
Engineering Students & Trainees: Those learning to identify material failures using datasets like NEU-CLS can use the “Beginner Mode” to receive guided, visual walkthroughs of specific defects like Crazing or Inclusion.

Quality Assurance (QA) Inspectors: Factory floor personnel who need immediate, AI-driven second opinions on surface defects during the manufacturing process.

Senior Metallurgical Engineers: Experts who require “Data-Dense” views, such as pixel intensity histograms and baseline comparisons, to make high-stakes “Accept/Reject” decisions.

Key Benefits
Eliminates “SaaS Bloat”: By using the “Kill the Dashboard” approach, users aren’t overwhelmed by unnecessary menus; the UI only generates the specific tools or charts needed for the current task.

Accelerates Learning: The A2UI (Agent-to-UI) protocol transforms a standard chat into a “Live Lab,” allowing students to interact with 3D models or heatmaps rather than just reading text.

High-Precision Analysis: By leveraging Google Vertex AI (Gemini 1.5 Pro), the system provides structured JSON data (bounding boxes and confidence scores) that ensures industrial-grade accuracy.

Scalable & Cost-Effective: Utilizing Google Cloud credits and serverless architecture (GCP Cloud Functions) allows the system to scale without the overhead of maintaining local heavy infrastructure like Docker.

N/A

AI Tinkerers CopilotKit Google Cloud Google DeepMind Jiva Infotech Inc Vertex AI

Github

Summarizing URL...

project website

Summarizing URL...