IBM · India Software Lab · Apr 2024

Eureka — ISL Cloud Hackathon conceptual design.

A 3-day cross-functional innovation sprint at IBM India Software Lab — competing to solve how IBM's AI Assistant can be optimised to retrieve location-based cloud resources accurately. Delivered competitive analysis, research, use-case mapping, and a conceptual design prototype.

AI UX Rapid Prototyping Hackathon Location Intelligence IBM AI Assistant

Duration

3 Days · 17–19 Apr 2024

Host

IBM India Software Lab

Format

Cross-functional teams

Credential

Eureka-ISL Cloud Hackathon

Context & Starting Point

How it started

IBM's internal Eureka Hackathon at the India Software Lab is a competitive innovation programme where cross-functional teams tackle real product problems in 3 days. In April 2024, the challenge put to participants was directly tied to IBM Cloud Location Management — a problem I was already embedded in as part of the Location Management UX research workstream.

Problem Statement: How can we optimise and train data for the IBM AI Assistant to retrieve location-based resources accurately?

This wasn't a theoretical prompt — it reflected a genuine gap in how IBM Cloud surfaced location-contextual information through its AI layer. Resources existed in specific regions, zones, and data centres, but the AI Assistant had no reliable way to answer location-specific queries like "show me my resources in Dallas" or "what services are available in eu-de-1?"

IBM Digital Credential

Eureka-ISL Cloud Hackathon badge

Participation in the Eureka ISL Cloud Hackathon is recognised as an IBM Digital Credential. The badge below shows all 8 IBM Digital Credentials earned across cloud, platform, product, and design domains.

IBM Digital Credentials — Product-led Certification · Eureka-ISL Cloud Hackathon · IBM Cloud Technical Advocate Concepts V2 · Enterprise Design Thinking for Sustainability · Enterprise Design Thinking Team Essentials for AI · Co-Creator · Practitioner · IBM Consulting Way Habits

Sprint Work

What was delivered in 3 days

Day 1

Competitive Analysis & Research

Benchmarked how AWS, GCP, and Azure surface location-contextual information through their AI assistants and search layers. Identified patterns — regional filter suggestions, location-aware resource queries, contextual tooltips — and mapped them to IBM Cloud's current gaps.

Day 2

Use-case Mapping & Data Model

Defined the key use cases: (1) "Show resources in [region]" query, (2) "What services are available in [zone]?" lookup, (3) Location filter persistence across AI session context. Mapped how location data (Geography → Country → Metro → Region → Zone) needed to be structured for AI training.

Day 3

Conceptual Design Prototype

Designed a conceptual prototype showing the IBM AI Assistant responding to location-based queries with structured, accurate results — using the unified location hierarchy model as the retrieval backbone. Prototype demonstrated how training data should be structured and how the assistant should surface, confirm, and persist location context across a conversation.

Connection to Broader Work

How this fed into Location Management

The Eureka Hackathon was the catalyst for the full Location Management research programme. The problem statement surfaced at the hackathon directly informed the Phase 1 research brief — confirming that the location inconsistency issue wasn't just a UI problem, but a data and AI retrieval problem too.

The competitive analysis work done in 3 days during the hackathon became the foundation for the formal competitive analysis phase, and the use-case mapping directly shaped the E2E user journey research scope that followed over the next 14 months.

"The hackathon confirmed something we suspected — the location problem in IBM Cloud was systemic. It wasn't just about which dropdown to use. It was about whether the platform even knew where things were."

Sprint Retrospective — Eureka Hackathon, Apr 2024

Public Cloud Solution Tool

Verizon AI