ENGAGE · TEAM
Principal · AI Data, Interfaces and Knowledge Stores
Experience21+ years
Arun is Bugni Labs' principal for AI Data, Interfaces, and Knowledge Stores, the polyglot engineering discipline that turns raw enterprise data into structures AI can reason over, and that turns AI outputs into governed, audit-ready surfaces. He operates as both architect and hands-on engineer: designing backend APIs, microservices, and knowledge substrate patterns, and building them in production. 21+ years of engineering across enterprise systems, full-stack polyglot across Java, Python, TypeScript, and Node.js.
The data and event architecture underneath financial systems. The recurring threads are event-driven design for payments and screening, CQRS and event sourcing in core banking, ISO 20022 migration, and incremental routes out of a legacy estate. The through-line is that AI is only ever as good as the data structures and interfaces beneath it.
10 pieces · 1 Perspective · 3 Field Notes · 6 Guides
Chain-of-thought prompting reduced the score variance by more than half, making LLM-based conformance scoring more consistent.
Read field noteLocal inference on 16GB developer laptops missed the performance bar; the hosted control it meant to replace matched its accuracy for pennies.
Read field noteA governed roadmap for modernising legacy banking systems with event-driven coordination, bounded agents, and reversible delivery.
Read guideHow event-driven architecture gives regulated financial platforms the real-time coordination, evidence, and controls needed for agentic work.
Read guideEvent-driven architecture in payments screening is the use of durable events, explicit domain contracts, and governed workflow orchestration to screen payments and related customer activity without turning every check into a blocking call chain. A payment platform publishes facts about what has happened. Screening, fraud, case management, operations, and reporting services react to those facts through controlled contracts.
Read guideFinancial systems architecture is the set of domain boundaries, data contracts, integration patterns, runtime controls, and operating responsibilities that allow a financial institution to change safely. It connects customer channels, products, payments, risk controls, ledgers, data platforms, AI services, and external networks without making every change a coordinated release across the bank.
Read guide## What are CQRS and event sourcing in financial services?
Read guideISO 20022 is an international standard for defining financial messages with a shared business model and structured data. For banks, migration is not simply a file-format conversion. It is the work of preserving the meaning, lineage, validation, and operational use of payment data as messages move between channels, payment engines, sanctions controls, ledgers, reconciliation, investigations, reporting, and AI-assisted operations.
Read guideEnterprise AI in financial services is being sold as a model problem. Almost every time we look closely, it is a data problem. Most banks are evaluating models without seriously evaluating where their data lives, what shape it is in, and whether the AI they are buying can do anything useful with it.
Read perspectiveA credit decisioning platform reached production by narrowing domains, using event contracts, and making every decision explainable.
Read field note