Agentic AI for private markets — built by someone who's run the operations.

The right solution for your situation — vendor, extend, or build. Assessed honestly, with 21 years of private markets and banking depth. Rigorous analysis of one vendor proposal saved over $500K at a leading PE firm.

$350B+

wires since inception

$500K+

saved — one vendor decision

0%→99%+

straight-through processing

21 yrs

BFSI · banks & PE

Featured case study

$350B in wires, 0% to 99%+ STP — rigorous analysis, right architecture A top PE firm · Treasury & Middle Office · Kyriba + SWIFT + GS Transaction Banking

Wire operations ran on fax and bank portals. A vendor came with a proposal — we evaluated it rigorously and declined. The analysis showed the firm already owned the right rails. Purpose-built automation on existing infrastructure achieved 99%+ STP, absorbed ~18x volume growth without headcount, and saved over $500K in vendor spend.

Read the case study →

Selected projects

Moneta

Next.js 14 · Hono · SQLite · TanStack Query

Local-only personal finance platform tracking six brokerages with TWR/MWR/Sharpe.

FlowDeck

Rust · flowengine · JSON filter DSL

Replaced a paid options-flow SaaS with ~800 lines of purpose-built Rust.

SanthoshIAS

Rust · priority-chain DataResolver

Financial data gateway routing across six market-data providers.

APEX / AlphaEdge

Bayesian expectancy · state machines

Options analytics platform with market-character classification, v11.

All projects →

Writing

Practitioner notes on fund technology and agentic AI — first pieces publishing soon.

About the writing →