DHANESHVAR MAHA@Dhaneshvar

Problem: Enterprises run an estimated 220 billion lines of COBOL in production, much of it on IBM mainframes. COBOL expertise is retiring faster than it can be replaced, and manual modernization takes months per program, costs millions, and risks breaking business-critical logic. Existing tools fail because they can't parse COBOL's fixed-format syntax and dialect quirks, can't extract business logic from poorly documented code, can't preserve JCL workflows and file interactions, and can't scale manual review across entire application portfolios. Our target users are mainframe operations teams, enterprise architects, and IT modernization leads—especially in finance and insurance—who need automated, trustworthy conversion, not just a scaffold generator. Solution: LM Modernizer is an end-to-end pipeline built around IBM watsonx.ai reasoning models. A custom ANTLR4 grammar parses both fixed- and free-format COBOL (including COBOL-85 and GnuCOBOL dialects), auto-detecting the format and building an AST. This is normalized into a language-agnostic intermediate representation capturing paragraphs, data divisions, file operations, and control flow. From there, a multi-agent system takes over: a Planner agent analyzes the IR and produces a migration plan with function dependencies and refactoring hints; an Executor agent generates Python/Java code, iterating across three rounds using Watsonx's extended reasoning to converge on faithful translations; a Critic agent validates the output—checking business-logic preservation, data type/range handling, file I/O fidelity, and exception coverage. A web dashboard shows migration progress, complexity hotspots, and risk scores, with every generated function traceable back to its original COBOL source. Interaction model: Users upload COBOL files, the platform parses and analyzes them, the AI planner generates a migration plan, users watch real-time progress on the dashboard, and then download modernized Python/Java code with full traceability to the source. What makes it unique: It's a first-of-its-kind grammar-based COBOL parser with self-healing format detection (fixing indentation misclassification automatically) and case-insensitive resolution of legacy naming patterns. The multi-round executor combines generation with reasoning-and-critique rather than one-shot translation. The IR layer bridges COBOL-specific idioms (CICS/IMS transactions, DB2/IMS file I/O, JCL batch jobs) to modern equivalents (async calls, ORM frameworks, Kubernetes/Docker manifests), and the architecture is modular enough to slot directly into watsonx Orchestrate as a Parser → Planner → Executor → Critic workflow. Effectiveness: By automating the labor-intensive parts of modernization while keeping a human-in-the-loop review step, LM Modernizer cuts manual effort by an estimated 80-90% per program, turning portfolio-wide modernization projects from a multi-year effort into one measured in months. Watsonx's reasoning catches edge cases that pattern-matching tools miss, and the traceability layer gives compliance and audit teams the paper trail they need to trust AI-generated code before it goes into production.

The lack of user-friendly tools creates barriers to independent citizen to access services on their own, leading to increasing administrative burdens and costs. We propose a Gen AI-powered platform using Grok models to simplify navigation and automate tasks on government websites. This empowers citizens to access services independently, reducing reliance on personnel and enhancing accessibility. User initiates a query with website and information request. Gen AI agent processes the request and accesses the target website in front of user and provide the screenshot of their flow to reach Target state. Gen AI extracts relevant data and generates a concise and understandable response about the given website

Support teams often struggle to manage increasing customer queries, leading to delayed responses and incorrect issue resolutions. IBM Diagnostics Gen AI Assistant addresses this challenge by automating troubleshooting processes, enabling faster and more accurate solutions. The system analyzes queries, identifies root causes, and suggests solutions, reducing manual intervention and increasing efficiency. This automation allows support teams to handle larger volumes of issues, scale operations, and improve overall response times. Ultimately, IBM's AI Assistant enhances customer satisfaction by ensuring issues are resolved quickly and correctly, leading to a more streamlined and effective support experience.

This project focuses on providing farmers in underserved communities with an intelligent AI-powered chatbot that addresses both general and complex queries related to plant care and farming. The system uses a Strategic agent approach: A local model (LLaMA 3.2 3B) running on an edge device to answer general, everyday plant care questions. A larger, cloud-hosted model (LLaMA 3.1 70 B) accessible via API for handling more complex or specialized queries. An intelligent agent is designed to assess the complexity of each query and assign a score. Based on the score, the agent dynamically routes the query to the appropriate model, optimizing for response time, computational resources, and accuracy. Farmers will interact with the system through a simple web or mobile chat interface (powered by Streamlit), enabling them to ask questions about plant health, harvest cycles, and more.

💻 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

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.

Leverage Orchestrate’s digital skills to design solutions that automate repetitive tasks, orchestrate workflows across tools, and empower employees to focus on high-value work. ⏳ Complete your project in just 48 hours and showcase your innovation. 🚀 Gain hands-on experience building with IBM watsonx Orchestrate. 💡 Develop practical AI agents that reshape the way teams and organizations operate. ⭐ Learn directly from experts who will support you throughout the event. 🤝 Team up or work independently to bring your vision to life. 💰 Prizes: $10,000 🧑🏻‍💻 Sign up before the Kick-Off Stream and start shaping the future with IBM watsonx Orchestrate!