Senior Continuous Improvement Engineer & Innovation Lead
CI, Engineering, Data, Legal, Cybersecurity, Ops
Azure OpenAI, RAG Architecture, RFP Process, ROI Modeling
AI as an enabler for existing bottlenecks—proving value before scaling
As Geidea scales its digital footprint, the volume of merchant interactions and internal operational complexity has grown exponentially. While traditional automation (RPA) handled structured tasks, complex queries and unstructured data remained a bottleneck.
The organization needed to move beyond "basic automation" to "cognitive intelligence"—scaling support and operations without linear headcount growth.
The organization faced a dual challenge requiring strategic clarity before tactical execution.
Human agents were bogged down by repetitive queries and manual data entry tasks like invoice logging—work that consumed resources without adding strategic value.
Institutional knowledge was scattered across disjointed repositories, making it difficult for employees to find answers quickly and serve merchants effectively.
Pressure existed to "use AI," but no clear roadmap defined where it would drive actual ROI versus being a costly gimmick.
"The opportunity: launch an AI Capability Activation Program to identify, validate, and architect high-impact Generative AI use cases across the customer and employee lifecycle."
Instead of rushing to full-scale digital overhaul, I devised a "Proof of Value" strategy—leveraging AI as a specific enabler for existing bottlenecks. These initial pilots serve as "Low Hanging Fruit" designed to secure ROI and governance maturity needed to launch a formal AI Center of Excellence.
AI-driven Voice Bot for Customer Care. Handles Tier-1 queries, verifies identity, creates tickets with smart escalation—deflecting volume from human agents.
Merchant Portal chatbot accessing live user data (transactions, terminal status) to answer specific queries like "Why was my settlement rejected?"
GenAI automation for unstructured tasks: feedback categorization, request routing, and OCR for invoice data extraction.
RAG-powered engine combining scattered knowledge bases (Sales, Tech, Ops) into one queryable interface—an internal "Co-pilot" for employees.
Since this technology is new to the organization, the process focused on Governance, Selection, and Readiness—building the foundation before deployment.
Modeled potential ROI for each use case using operational metrics (e.g., Deflection Rate × Cost per Ticket = Savings).
"We established AI guardrails before AI capabilities—ensuring governance maturity matches technical ambition."
I structured execution into a multi-phase roadmap to manage risk and build organizational confidence progressively.
Cleaning knowledge base data and establishing the AI Governance Framework. Building the infrastructure before deploying capabilities.
Launching the Internal Knowledge System first to test accuracy without risking customer trust. Employees as first users.
Deploying GenAI Automation (OCR/Tagging) with Human-in-the-Loop review. Building confidence through supervised outputs.
Deploying Voice Bot and App Assistant once confidence thresholds are met. Full customer-facing AI activation.
Although in pre-production, the strategic groundwork has delivered immediate organizational value and cleared the path for execution.
"We transformed AI from organizational pressure into strategic clarity—proving value on paper before proving it in production."