Who we are
We build the infrastructure that lets any desk offer a structure made for its client.
OpenArrow puts financial engineering within reach of every desk, so investors can access solutions designed around their needs, not just their ticket size.
OpenArrow is an AI-native platform for structured products and options analytics. Wealth desks, issuers, fund managers and fintech platforms use it to price, stress-test and document structures in one run, on one deterministic engine, inside the tools their teams already use.
We are vertical infrastructure for finance: the pricing and documentation layer that an assistant, a platform or a structurer calls when a number has to be right.
- Based in
- India
- In production
- Since June 2026
- Underlyings
- NIFTY · Gold
- Engine
- arrowquant
- Access
- Chat · MCP · API
- For
- Desks · Issuers · Platforms
Our story
Started on a structuring desk.
OpenArrow began with a question from years of pricing and structuring derivatives: why does a product built for one client cost so much more to make than one built for a thousand? The answer was the manufacturing, not the maths. So we built the engine and the workflow around it.
- Jun 2026
Live in production
Structured products on the Nifty 50 Index, from chat, MCP tools and the Structures shelf.
- Aug 2026
Stanford Seed Spark
Selected for Seed Spark 13 South Asia, the Stanford Graduate School of Business programme for early-stage founders.
- Sep 2026
arrowquant engine and gold
A new engine behind every surface, so the quote, the shelf and the KID agree. Gold added as an underlying.
- Now
Desks and classrooms
In conversation with desks and issuers, and offering academic licences to universities that teach structuring.
The team
Built by people who have sat on the desk.

Founder
Kannan Singaravelu
“It started on a structuring desk at Kotak, wiring in Sobol sequence generators and building Excel add-ins just to price one structured note. Later at Bloomberg, working with derivatives desks worldwide, I kept seeing the same thing. The maths was never the hard part. It's everything around it.”
Previously Kotak, structuring; Bloomberg, equity derivatives; Mirae Asset Sharekhan. CQF instructor with Fitch Learning. Creator of quantmod, an open-source Python library for quantitative research. Stanford Seed Spark 13 South Asia, Stanford Graduate School of Business.
Advisory
Technology and product architecture, from people who have run it at scale.
Rajesuwer Singaravelu
Financial technology · Product architecture
Over twenty years in financial technology and product architecture. Formerly at Credit Suisse, USA.
LinkedInRavikumar Govindaraj
Cloud infrastructure · Automation
Over twenty years in cloud infrastructure and automation. Solutions Architect at Broadcom.
LinkedInHow we build
The model talks. The engine computes.
No figure on a quote, term sheet or KID is written by a language model.
Every number has an owner.
Each engine call is logged against the key that made it, so any figure can be traced back.
Desk first, software second.
We build the workflow a structurer already runs, and take the hand-offs out of it.
Talk to the people who built it.
Questions on the engine, a pilot, or working with us.