2017 — 2019
Reconciliation Analyst
State Street Global Advisory
Reconciled 100+ managed pension fund portfolios across equities, derivatives, FX, and post-trade settlement.
Abhilash
Kumar Panda
Senior Data Scientist · GenAI · LLMOps
I build production AI for financial markets — by people who've read the documents it processes.
9 yrs experience · £600K+ business value · 55 FTEs automated · 130K+ documents processed
01 / The Journey
From the desk to the model layer
Three chapters. Each one taught me something the next chapter needed. The domain came first; the models came afterwards — and that's exactly why they worked.
2019 — 2022
Data Scientist
London Stock Exchange Group
Moved from operations to automation — building NLP pipelines and ML systems for financial documents.
2022 — Present
Senior Data Scientist
London Stock Exchange Group
Leading GenAI and agentic systems that deliver £600K+ in production value across global workflows.
02 / Principles
What I believe
Domain-first
Models don't fail on syntax. They fail on context — the settlement logic, the regulatory nuance, the edge case that only shows up in month three.
Compliance-aware
Every production system is source-referenced, auditable, and explainable. In finance, trust is the real deliverable.
Outcome-driven
I optimise for business value, not leaderboard scores. £600K+ delivered and 55 FTEs automated are the only metrics that matter.
03 / Proof
Production work that proves the thesis
GenAI Financial Data Extraction Pipeline
£350K+ saved · 50 FTE reduction
Korean Regulatory Filing Extraction
15 days → 40 minutes · 130K+ docs
04 / What's Next
Bring the unfair advantage to your team
Nine years in. I'm looking for senior IC and lead roles at the intersection of financial services and production AI — places where models are the product, not the slide deck.
If you need AI that understands settlement logic, regulatory nuance, and analyst workflows, let's talk.