
3
2
Pakistan
1 year of experience
I’m an experienced full-stack developer with a strong command of Next.js, Firebase, and n8n for building powerful, automated web applications. My expertise extends to Python SDKs and Gemini CLI, enabling me to integrate advanced AI capabilities and automation into real-world solutions. Over the years, I’ve developed and deployed modern, high-performance websites and intelligent systems that deliver both speed and scalability. My deep understanding of front-end and back-end technologies allows me to handle everything from UI/UX design to complex backend integrations with ease. I’m passionate about creating innovative digital products that combine clean design, cutting-edge technology, and seamless user experience. I continuously explore new tools and frameworks to stay ahead in the evolving tech landscape.

Our project, the GSCR Agentic Control Center (Geopolitical & Supply Chain Risk), is a fully automated solution designed to protect enterprises from high-impact global disruptions. The core problem is the multi-million dollar loss incurred when companies react too slowly to sudden events like port closures or regional conflicts. Our system transforms this manual reaction process into instant, orchestrated resilience. The GSCR Agent is built on IBM watsonx Orchestrate and operates with a sophisticated Agentic Framework. When a risk is detected, the Agent does not wait for human intervention. Instead, it autonomously initiates a detailed workflow: 1: Detection & Analysis: The Agent calls its Digital Skills (APIs) to instantly check the status of critical supply ports and identify all impacted inventory (SKUs). 2: LLM Reflection and Decision: It uses LLM Reflection against a pre-defined Mitigation Strategy Knowledge Base (business policy) to rank available alternative logistics routes and determine the single best option based on the trade-off between cost and speed. 3: Human Handoff: The Agent presents this pre-vetted, optimal solution on our Dynamic UI for necessary one-click human approval (Human-in-the-Loop). 4: Autonomous Execution: Upon approval, the Agent calls its final Digital Skill to execute the route change in the SCM system immediately. By leveraging watsonx Orchestrate's power, the GSCR Agent reduces the time required to stabilize the supply chain from weeks to mere minutes, delivering a massive ROI and unparalleled operational security.
23 Nov 2025

Opus-ClaimGuard is a robust, end-to-end workflow designed specifically for the Applied AI Challenge to bring trust and efficiency back to insurance claims processing. As a solo developer, Anum Munir, I focused on demonstrating the highest level of platform efficiency and auditability achievable. The Solution: This project features a unique hybrid decision engine built on the Opus platform. It is capable of combining strict, rule-based policy checks (e.g., auto-rejecting claims with an expired license) with deep Generative AI analysis. Winning Feature: Multimodal Auditability. The system handles multimodal input by simultaneously analyzing the claim document (for extracted text data) and the damage photo (visual evidence). This allows the AI to perform a crucial consistency check and assign a Fraud Probability Score. For high-risk cases, the system generates a full Audit Log that includes the AI's Rationale (e.g., 'Minor damage does not match the $15,000 claim request'). This solves the 'black box' problem and provides complete $\text{Intake} \to \text{Deliver}$ traceability, which is essential for regulated industries. This project proves that one developer can build powerful, auditable, and enterprise-ready solutions.
19 Nov 2025