Mahnoor Sarfraz@mahnoor_sarfraz773

Customer Success Managers lose 20 minutes after every call re-listening for details, and churn signals — "we're evaluating a competitor," "our sponsor left" — get buried in a 40-minute transcript. AI Customer Success Copilot fixes this with a real pipeline, not a single prompt wrapper: a call recording is transcribed via Speechmatics, then passed to an LLM (AI/ML API / Featherless AI) that extracts a churn-risk score with the exact quote that triggered it, a call summary, key concerns, and commitments made. From that structured data, the app drafts a follow-up email and a task list — nothing sends automatically. The CSM reviews everything on one screen, edits inline, and approves with a single click before it's finalized. Approved calls persist to an account history and a team-wide risk dashboard, so a CS lead can spot which accounts need attention this week — not after the renewal is already lost.

Our project is an AI-powered coding assistant called *Code Copilot*, designed to help programmers improve and understand their code more effectively. It is built using Python, Gradio for the user interface, and Blackbox AI (GPT-4) to provide intelligent coding responses. When a user pastes a piece of code or asks a coding question, the system sends that input to the AI model to generate helpful advice and explanations. At the same time, the app uses Python’s ast (Abstract Syntax Tree) module to deeply analyze the structure of the code by identifying how many functions, loops, and conditionals it has, and how deeply the code is nested. Based on this analysis, the app automatically generates suggestions such as simplifying nested logic, breaking large functions into smaller ones, or replacing basic loops with more efficient techniques like list comprehensions. Everything—including the AI response, code pattern analysis, and smart suggestions—is displayed in a clean, user-friendly interface. This makes Code Copilot a powerful tool for both beginner and experienced developers to write cleaner, more efficient Python code with real-time feedback.

AI code Commentor and Documentation Generator aims to remove headache to remove writing comments, documentation for developers which is real pain for some beginner level developers fellow, we come with idea where a user interface only requires a piece of code and AI will automatically provide comments, documentation, provides improvements if any and gives the time complexity. with more than 90% success rate, we present our project to lablab with plans to broaden the scope of project in future. we are using Trae.ai as our IDE, llma from novita ai and fast api to handle api request also serving a static html page to provide interface to users. we provide simple user interface, to generate comments and documentation at ease.

💻 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

🌎 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 | 📅 August 3–10, 2026| 💻 Powered by native.builder The official native.builder hackathon, hosted by NativelyAI and lablab.ai. Join founders, developers, designers, and builders to create real AI-native applications powered by native.builder: from idea to deployed product, fully in the cloud. 💻 This hackathon is fully online. No travel required.