Finpilot Raises $10M Series A for Investment Office AI

In February 2024, Madrona's original investment thesis for Finpilot centered on helping financial analysts find information and put it to work. The investor described a chat tool alongside plans for agents, placing the young company close to the researcher's desk and the sources on the screen.

The business now reaches further into the investment office. Seattle-based Finpilot announced a $10M Series A on November 4, 2026, led by Lobby Capital with Madrona participating, to expand its product and team. Its customers manage institutional capital, and its software connects the information and recurring tasks surrounding those portfolios.

That progression gives the financing a practical meaning for institutional operators. A research tool helps someone answer a question; an operating platform takes on a continuing obligation to keep the office's information usable as documents, people and decisions change. Finpilot is financing more of that continuing work.

A financing for a longer institutional relationship

CEO and co-founder Lakshay Chauhan and co-founder John Alberg, the company's executive chairman, bring a relationship formed around investment technology. Madrona's early account describes Chauhan working on quantitative-investing technology at Euclidean Technologies, which Alberg co-founded. Their shared background helps explain why financial information was the starting material for the company.

GeekWire's current reporting places Finpilot's founding in 2023 and reports approximately $14.5M raised in total, including the latest round. It also identifies a $4M seed financing in 2024. The $10M announcement is the new Series A, while the larger figure describes cumulative financing; valuation and additional transaction terms were not disclosed in the sources reviewed.

The company says the new capital will support hiring and product expansion. Its website identifies institutional investment users including Seattle Children's, TIFF and the Wyoming Retirement System. These are company-disclosed relationships, placing the product inside organizations where information may have to serve several committees, mandates and reporting calendars at once.

For a supplier, that creates an unusually detailed customer assignment. An investment office already has ways to define a fund, approve a number and remember a manager conversation. Taking on recurring work means learning those practices closely enough that the next report still belongs to the institution receiving it.

Connecting records without starting the office over

Finpilot's Connected Data documentation describes connecting existing systems while leaving them authoritative for the information they own. The platform links references to investments, organizations, people and accounts across documents and applications. Where a possible match or conflicting value needs attention, that uncertainty remains available for review.

Consider the ordinary difficulty implied by that design: an investment can appear under different references in a manager document, a portfolio system and a meeting note. Joining those references changes what a team can ask across its information. Joining the wrong ones would give the office a coherent-looking account of the wrong relationship, which makes visible uncertainty part of the useful product.

Keeping the underlying system's authority also changes the adoption conversation. The institution can evaluate a new layer of work around existing records without first agreeing to replace every application involved. That gives Finpilot a route into the office while increasing the importance of understanding how responsibility is already distributed among its systems.

The commercial consequence follows from that architecture: a deployment carries the customer's definitions and history with it. More documents alone would leave the institution with more material to search. Connected references can make past decisions useful in the next assignment, provided the team can still inspect which source supplied the information.

The approval point is part of the workflow

The company's agent workflow description shows monitoring, reporting and meeting follow-through configured around permitted sources and review points. In its illustrative manager-monitoring workflow, a material change stops for investment-team judgment. These are demonstrations of intended behavior, rather than independently measured customer results.

An approval point gives the product a place in the institution's division of labor. An agent can prepare an update while the investment team decides whether the underlying change belongs in committee materials. Moving that responsibility into a configured process requires the office to identify who may decide and what evidence should accompany the handoff.

That is consequential design work for software sold as automation. The organization has to make some of its own practices explicit before the agent can reliably advance them. Finpilot's opportunity includes helping those practices survive the next report, the next meeting and the next person who needs to understand the record.

Finpilot's security documentation says institutional information is separated by customer, access is governed by roles and permissions, and customer data is excluded from model training under its approved policy and applicable provider contracts. It also states SOC 2 Type II assurance, with the report available during diligence. Buyers still need that diligence to assess the controls against their own requirements; the public policy describes the vendor's position rather than supplying an independent assessment here.

Hiring for the work between a product and an office

Finpilot's current careers page lists Seattle roles across software engineering, product, marketing, sales, client success and business operations. The mix includes an AI Solutions Specialist in implementation and client success. That position sits close to the practical job created when a customer asks software to understand how its investment team works.

The surrounding market opportunity is specific: institutional offices have recurring information work and established ways to approve it. Finpilot is trying to make a connected record useful repeatedly, across research, monitoring and reporting. A buyer can examine that proposition by following a single workflow from incoming evidence to the person responsible for the result.

For the company, growth brings another set of institutional habits into the product relationship. Each new office has its own language for investments and its own path from manager communication to committee judgment. The Series A gives Finpilot more people and product capacity to accompany that path, with the next implementation beginning inside the customer's existing records.