Ishank Gupta@Ishank

SayQL is a voice-first AI data analyst that helps business owners understand their data without writing SQL or building dashboards. A user can simply ask questions like, "How many returns did we have last week by city?" or "Why did returns increase last month?" and SayQL turns the conversation into a clear answer with tables, charts, SQL, and supporting evidence. Built around AssemblyAI's Voice Agent capabilities, SayQL supports natural voice interaction, real-time tool calling, and interruption-aware conversations. When the agent needs data, it can call backend tools to inspect the business schema, resolve the relevant metric, execute a guarded read-only SQL query, and use the results to generate the response. The key problem we solve is not just SQL generation. Business metrics often have different meanings across companies. "Net revenue," "active customer," or "return rate" cannot safely be guessed by an LLM. SayQL therefore uses a semantic metric layer with explicit business definitions. The agent retrieves the relevant definition before querying the data, and when a term is ambiguous, it asks the user instead of making an assumption. Every answer is designed to be transparent: users can see the metric definition used, the generated SQL, the result data, and the visual explanation behind the answer. This makes SayQL useful not only for asking questions, but for understanding and trusting the answer. Our vision is to make analytics feel as simple as having a conversation with a reliable data analyst, starting with small D2C businesses and evolving into a broader AI-powered business intelligence platform.

AI Voice + Vision Emergency Assistant is a multimodal AI system designed to help users communicate and understand emergency situations using voice, images, and contextual information. The application combines speech-to-text, computer vision, and an emergency reasoning agent into one end-to-end workflow. A user can describe an emergency through voice and optionally provide an image of the situation. The voice module converts speech into text, while the vision module analyzes the image for potential hazards and relevant objects. These inputs are then combined by the emergency agent to determine a possible emergency type, severity level, confidence score, reasoning, and recommended safety actions. The project uses a React frontend and FastAPI backend to connect the different AI components. The backend provides an API for multimodal analysis, while the frontend presents the emergency assessment in an accessible dashboard. The goal is to demonstrate how modern AI infrastructure can combine voice, vision, and agentic reasoning into a practical real-world safety application. The system is designed as an emergency-assistance prototype and is not intended to replace professional emergency services or human judgment.

Deal Desk is a Band-native multi-agent system that automates enterprise proposal generation from customer RFPs. Enterprise sales teams often need pricing, technical, and legal stakeholders to collaborate before responding to a customer. This process is slow, involves multiple handoffs, and creates bottlenecks that delay deals. Deal Desk solves this problem using a self-assembling team of AI agents running inside Band. When an RFP is submitted, a Coordinator agent analyzes the requirements and dynamically recruits only the specialists needed for that specific deal. A simple renewal may require only a Pricing Specialist, while a complex enterprise opportunity may recruit Pricing, Technical, and Legal Specialists. Each specialist creates its section of the proposal. A Reviewer agent then evaluates the output, identifies weaknesses, and requests revisions until quality standards are met. The system also includes human-in-the-loop guardrails. If a proposed discount exceeds 20%, the workflow pauses and requests human approval before proceeding. All collaboration happens natively inside Band through agent discovery, recruitment, messaging, review loops, and human participation in a shared room. Key features: 1. Dynamic specialist recruitment 2. Automated review and revision loop 3. Human approval workflows 4. Band-native orchestration 5. Enterprise proposal generation Deal Desk demonstrates how AI teams can transform a multi-day internal workflow into a process completed in minutes.

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

🌎 Hybrid Hackathon | 💻 Online Build: 3–8 November 2026 | On-site Phase: 8 November 2026 📍 On-site Venue: Pierpont Place 163 W Pierpont Ave, Salt Lake City, UT 84101, USA ✨ Please note that travel and accommodation are not covered.

🌎 Hybrid Hackathon | 💻 Online Build: September 10–16, 2026 | 📍 On-site: September 15–17, 2026 The official hackathon of AI Infra Summit — co-organized by Kisaco Research and lablab.ai. Build production-grade AI applications on the modern AI infrastructure stack, alongside 8,000+ attendees, researchers, founders, and infra leaders. 📍 On-site Venue (September 15–17): Santa Clara Convention Center 5001 Great America Pkwy, Santa Clara, CA 📍 On-site participation is by invitation only. Travel and accommodation expenses will not be covered.

🌎 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.