A Seattle data company built on a stubborn idea: the customer isn't a record. They're a person scattered across a dozen systems - and only machine learning can put them back together.
Every consumer brand collects more data about you than it can actually use. You are an email address in one system, a phone number in a loyalty database, a billing address at checkout, a device ID in an app, and a misspelled name at a call center. To the company, that scatter looks like a dozen strangers. To you, it is just one Tuesday. Closing that gap - turning fragments into a single, trustworthy customer - is the entire business of Amperity.
Founded in Seattle in 2016 by Kabir Shahani and Derek Slager, Amperity sells an AI-powered customer data platform (CDP) to some of the largest consumer brands in the world. Airlines, retailers, hotel chains, and sports franchises use it to answer a deceptively simple question: who is this customer, really? The answer sits underneath loyalty programs, personalized offers, and increasingly the AI agents brands now want to point at their customers.
The founders had seen the problem before. Enterprises were pouring years of engineering into brittle rule-based systems that tried to "fix" data by hand. Every new channel or identifier added complexity, and the harder teams pushed, the faster the systems broke. Shahani and Slager - who had already built and sold a company called Appature together - decided the mess was not a data-hygiene chore. It was a hard computer science problem, and they would treat it like one.
Figures compiled from public funding announcements and press reporting. Revenue and headcount are approximate.
At the center of Amperity is Stitch, a patented identity resolution engine. Instead of relying on one shared key - which never exists cleanly across real-world data - Stitch combines deterministic matching (exact, rule-based links) with probabilistic matching (statistical likelihood that two records are the same human). Amperity has described running up to 45 different machine-learning algorithms to build a single identity graph.
The design has an unusual and genuinely useful property: it can admit mistakes. When an identity graph wrongly fuses two different people into one profile - an "over-merge" that quietly corrupts everything downstream - Stitch can decompose the record and pull them apart again. Most systems cannot even detect that error, let alone reverse it.
A simplified view of Amperity's identity resolution: scattered identifiers in, a single governed profile out - with no rigid schema required at the door.
Ingest data in almost any format - no upfront schema - and get clean, deduplicated profiles that span channels, devices, and identifiers.
Retail-proven predictive models, sharpened by the Custora acquisition, forecast customer lifetime value and who is likely to leave.
Build segments and push unified profiles to marketing, analytics, and BI tools so teams act on the same source of truth.
Secure data sharing with Snowflake and Databricks - bi-directional access without copying data or scheduling reverse ETL.
Role-based access, data governance, and privacy controls (SOC 2, GDPR-oriented) so identity work stays auditable.
Talk to your data in the terminal: tag it, resolve identities, and delegate customer-data engineering to an AI agent inside Databricks.
Amperity's clients are large B2C brands where the data mess is worst and the payoff for solving it is biggest. Public customers have included Alaska Airlines, DICK'S Sporting Goods, Wyndham Hotels & Resorts, Michaels, Cost Plus World Market, Kendra Scott, Tapestry, European Wax Center, KUIU, the Smithsonian, and even Seattle Sounders FC.
Its identity models are trained on billions of records drawn from more than 400 enterprise brands - the kind of accumulated pattern library that is difficult for a newcomer to replicate.
"The best AI is only as good as the data underneath it."
The CDP category is crowded - Salesforce Data Cloud, Adobe Real-Time CDP, Twilio Segment, Treasure Data, Tealium and others all compete for the same buyer. Amperity's wedge is narrower and deeper.
| Dimension | The common approach | Amperity's approach |
|---|---|---|
| Data intake | Rigid schema defined up front | Schema-free ingestion; the ML layer handles inconsistency |
| Matching | Deterministic rules or single-vendor IDs | Deterministic + probabilistic ML, up to 45 algorithms |
| Bad merges | Hard to detect, rarely reversible | Stitch can decompose incorrectly merged profiles |
| Where data lives | Copied into the vendor's walled garden | Runs in your lakehouse (Snowflake, Databricks) via Bridge |
| AI posture | Bolted-on features | Rebuilt AI-first; agents like Chuck Data on governed identity |
Comparison reflects Amperity's public positioning; individual competitor capabilities vary and evolve.
Amperity is a B2B enterprise software company. It sells subscription access to its platform, generally priced by data volume, usage, and the modules a brand turns on. That model rewards land-and-expand: a retailer might start with identity resolution, then add predictive analytics, then wire profiles into its warehouse.
Newer motions push the product closer to where enterprises already keep their data. The 2024 launch of the Customer Data Cloud and the Bridge integrations positioned Amperity as, in its own words, the world's first lakehouse-native CDP - a bet that the future of data tooling is "come to my data," not "send me your data."
In the market, Amperity sits at the infrastructure layer of martech: the identity foundation beneath marketing, analytics, and now AI agents. It is not the flashy consumer app; it is the layer those apps quietly depend on. That is a durable place to sit - and a demanding one, because when identity is wrong, everything built on top of it is wrong too.
The company reached unicorn status in 2021 after posting nearly 100% revenue growth the prior year. It has since been named a Leader in IDC MarketScape evaluations for both retail CDPs (2025) and AI-enabled CDPs (2026).
Shahani and Slager launch Amperity and raise a $9M Series A led by Madrona to solve identity with machine learning.
Backed by Tiger Global and Declaration Partners, the multi-patented Stitch engine scales across retail and travel.
Amperity folds retail-proven predictive customer lifetime value into the platform.
A $100M Series D led by HighSage Ventures values the company north of $1B; total funding reaches $187M.
Customer Data Cloud and Bridge launch - unified profiles shared into Snowflake and Databricks without reverse ETL.
The first AI agent for customer data engineering debuts at the Databricks Data + AI Summit.
Founders return as co-CEOs and restructure the company around AI; Amperity is named an IDC MarketScape Leader for AI-enabled CDPs.
University of Washington informatics grad and repeat entrepreneur. Co-founded Appature with Slager before Amperity. In 2026 he returned as co-CEO to lead the company's AI-first turn.
A software engineer by background, Slager drives Amperity's AI and identity vision - an AI-first platform grounded in accurate, governed customer data for agent-driven use cases at enterprise scale.
The flagship AI agent, "Chuck Data," is a data engineer you talk to in the terminal.
Shahani and Slager built and sold Appature together before starting Amperity.
Stitch may run up to 45 ML models to assemble a single identity graph.
Unlike most CDPs, it accepts messy data in any format and sorts it out.
Its models learn from billions of records across hundreds of enterprises.
Seattle Sounders FC sits on a customer list alongside airlines and hotels.
Links open YouTube. Specific video availability may change over time.
Amperity is an AI-powered customer data platform. It uses machine learning to unify fragmented customer records from many sources into a single accurate profile, then makes those profiles available for analytics, marketing, and AI.
It was founded in 2016 in Seattle by Kabir Shahani and Derek Slager, who had previously built and sold a company called Appature together.
About $187 million total. Its $100 million Series D in July 2021 valued the company at over $1 billion, making it a unicorn.
Stitch is Amperity's patented identity resolution engine. It combines deterministic and probabilistic matching using up to 45 machine-learning algorithms, and can even decompose records that were incorrectly merged.
It leads with ML-based identity resolution and schema-free ingestion, and runs lakehouse-native inside Snowflake and Databricks - sharing unified profiles without copying data or reverse ETL, rather than forcing data into its own walled garden.