hello

I'm Bharath

hello

I'm Bharath

building candela AI
at FINBOURNE

2008

Where I've been,

and what I've built.

2008 — Operations

2008

Operations
@ Northern Trust

Joined Northern Trust and rotated across Fund Accounting, Transfer Agency, Middle Office, Corporate Actions and Custody — building domain expertise across the full breadth of asset servicing operations.

2014 — Risk & Control

2014

Risk & Control
@ Northern Trust

Designed and implemented a data-driven first-line Risk & Control framework across Fund Accounting, Transfer Agency and Middle Office in India and Manila — reimagining how operational risk was identified, measured and managed.

2018 — Data & Analytics

2018

VP – Data Analytics
@ Northern Trust

Applied quantitative methods to surface the real issues hiding behind operational symptoms. Developed and ran an ML Bootcamp — teaching data science and machine learning to 260 people with zero prior coding or maths background.

2021 — Data Science

2021

VP – Data Science
@ Northern Trust

Led the build of an MLOps platform for Fund Accounting from the ground up. Brought machine learning and AI into core accounting workflows — using data and quantitative methods to solve problems that had previously only been seen through their symptoms.

2024 — Transformation

2024

SVP – Transformation
@ Northern Trust

Led the Transformation & Innovation office. Conceived and shipped internal and client-facing products that transformed Fund Accounting at scale. Two products featured in Scrip Spring Issue 2025 and WatersTechnology.

2025 — Present

2025

AI Lead
@ FINBOURNE Technology

Building Candela — FINBOURNE’s agentic AI platform that connects directly to live investment data via the Model Context Protocol. Agents that don’t just answer questions but execute real workflows: fund onboarding, compliance checks, portfolio analysis — against production financial data.

01

I build at the intersection where 17 years of financial domain depth meets engineering instinct — a combination rare enough that I’ve spent most of my career doing work that didn’t exist in a job description yet.

02

I don’t prototype. Candela — the agentic AI platform I’m building at FINBOURNE — executes real investment workflows against live production data. Fund onboarding, compliance checks, portfolio analysis: done by agents, not described by them.

03

I once taught machine learning to 260 people who had never written a line of code. That instinct — to make the genuinely complex feel inevitable — is how I build products too.


Study

Caltech — PG Diploma, DevOps  ·  UT Austin — PGP, AI & ML  ·  Uni. Ireland — IOB Banking  ·  Edith Cowan — MBA Finance (Gold Medal)  ·  JNTU — B.Tech, EEE

Stack

Python · C# / .NET · TypeScript · Rust · SQL  —  AWS Bedrock · MCP · Agentic systems · MLOps  —  Distributed systems · Data Vault 2.0 · Graph, relational & document DBs