About

We help software teams read their customers

Sova is the customer intelligence layer for software companies. We turn the scattered, unstructured things customers say into ranked decisions your team can act on and defend.

Why we exist

The answer is already in the feedback

Every software team is sitting on a pile of customer signal: tickets, call transcripts, reviews, NPS notes, cancellation reasons. Buried in it is the answer to what to build next, why customers leave, and which accounts are about to churn.

Nobody has the hours to read all of it. So teams decide from memory and the loudest voice in the room. Sova was founded in March 2022 to close that gap, by using language models to read everything and hand back ranked decisions with the exact evidence attached.

We are a small US team building Sova full time. Our founders and leadership are based in the United States, and we serve software companies worldwide.

Founded
March 2022
Stage
Pre-Seed
Headquarters
New York, NY
Team
14 people
Backing

Pre-Seed, $950,000, announced September 2022. Led by Ridgeline Capital, with Dana Whitfield and Samuel Reyes.

Team

The people building Sova

A small, senior team. Illustrative leadership for this launch, all building Sova full time.

NR

Nadia Roswell

Co-founder & CEO

Led product at two developer-tools companies. Started Sova after watching teams drown in customer feedback they could never read at scale.

ES

Elliot Shaw

Co-founder & CTO

Built retrieval and ranking systems for enterprise search. Owns Sova's model routing and evidence extraction.

MO

Maya Okafor

Head of Product

Ran research and roadmap at a Series B SaaS company. Turns scattered signal into decisions teams actually ship.

BR

Ben Rossi

Founding Engineer

Owns ingestion and the pipelines. Keeps every insight traceable back to the exact quote it came from.

Values

How we work

Evidence over opinion

No decision without proof. If Sova cannot point to the exact customer words behind a claim, the claim does not ship.

Read the frontier

The product improves as the models do. We stay model-agnostic so the best model for each job is always the one we use.

Customer language first

The answer is almost always already in what customers said. Our job is to read it at a scale no team can by hand.

Start turning customer signal into decisions

Connect one source free and see your first ranked decisions with the evidence behind them. No credit card.