Data, analytics & apps
— built to fit.

Polecat builds data, analytics, and application software for targeted problems — designed by thirty years of enterprise data experience, built by a fleet of AI agents that ship and test around the clock. All the output of a team. None of the meetings.

See what we've shipped Work with me →

Everything here is open source (GPL-3), local-first, and free to use without an account. Yes, really.

scroll
The idea

Most organizations don't need another platform.
They need their actual problems solved.

Enterprise software is priced for armies — armies to build it, armies to sell it, armies to sit through the rollout. So targeted problems go unsolved, because no one can justify the overhead for a tool that does one thing well.

Polecat works from the other direction. Thirty years of enterprise data experience sets the architecture, the standards, and the taste; AI agents do the building, testing, and shipping — continuously, in public, with every change gated and reversible. That changes the cost of software so much that "a precise tool for your exact problem" becomes the cheap option, not the extravagant one.

The goal is simple: elevate how organizations work, operate, and analyze — by streamlining and simplifying, not by adding another layer. Fewer moving parts, clearer data, tools people actually enjoy using.

7live apps
900+releases shipped
100%open source
0meetings held
Proof, not promises

The apps

Every tile is live and free — versions and ship dates come straight from each app's changelog. This is what agent-scale development looks like in production.

How we build

A factory that never sleeps

01 A shared platform, not eight snowflakes Every app runs on one open shell library — design tokens, navigation, theming, changelogs — vendored with integrity hashes. Improve it once, the whole fleet gets better.
02 Agents build; gates decide AI agents work in branches and open pull requests. Nothing ships until the app's own test gate passes on mobile and desktop — and if something bad ever lands, the site auto-reverts itself within a minute.
03 Shipped daily, in public Every change lands with an honest changelog entry, every repo is public, and every app tells you exactly what's new. Software you can watch being made.
04 Local-first, yours-first Apps work fully without an account; your data lives in your browser unless you connect a backend — including your own, because it's all GPL-3 and self-hostable end to end.

The same factory is available for your targeted problem. That's the pitch, and the site you're reading is the proof.

What I do

Independent data, analytics & AI consulting

I help organizations design, modernize, and improve data platforms, analytics environments, and AI-enabled solutions — moving fluidly from executive strategy to architecture to hands-on implementation and delivery. You don't have to choose between a strategist who can't build and a builder who can't see the business.

Data integration & pipelines Enterprise-scale ETL/ELT design and optimization, metadata-driven frameworks, migrations off legacy platforms (Informatica and friends), data quality, lineage, and the unglamorous performance work that makes it all hold.
Data, end to end The whole chain, ingestion to insight — with deep open-source roots: full-platform Pentaho depth (Data Integration, Server and clustering, Analyzer/Mondrian, Catalog, Quality, Optimizer), Mondrian and the wider open-source data stack, plus generative-AI extensions built for the platform itself.
Cloud & modern data platforms AWS, Azure, and GCP architecture; Kubernetes and containers; Snowflake, Databricks, BigQuery, Redshift; hybrid and multi-cloud designs; migrations sized honestly, with cost and operability treated as features.
Analytics & BI Semantic models, dimensional modeling, governed metrics, embedded and OEM analytics, dashboard modernization — turning business questions into data products people actually use.
AI & intelligent automation Generative AI for real data workflows: document AI, vector search and matching, prompt and evaluation frameworks, and agent-driven development — practiced daily, not read about. This entire site is the demo.
Architecture, strategy & delivery Assessments, modernization roadmaps, proofs of concept, build-vs-buy calls with no software quota behind them — plus the delivery leadership to land it: SOWs, estimation, escalations, and teams that ship.
Why this works

The unusual part isn't any single skill — it's operating credibly at four levels at once:

Executive strategy Enterprise architecture Hands-on implementation Delivery leadership

Assess → strategize → roadmap → prototype → implement → troubleshoot → enable your team to run it. Start anywhere in that chain; I've done every link of it for thirty years.

About

Kevin R. Haas

I've spent thirty years in enterprise data — from building predictive models in the '90s, through senior partner at Inquidia Consulting and director roles at FICO, to leading Pentaho Professional Services globally at Hitachi Vantara: large-scale data and analytics delivery, cloud migrations, GenAI product extensions, and the escalations nobody else wanted. Florida, then Northwestern, before all of it. Chicago ever since.

Polecat is what happens when all of that meets agent-scale development: an independent practice and an open-source software collective — of one — building the kind of precise, joyful tools I always wished every company could have. Enterprise-grade elegance, finally cost-effective for everyone from the smallest startup to the largest enterprise. The agents don't sleep. I occasionally do.

Connect

Connect with us

Want a key to a private app like the Model Server, curious about a build, or have a problem worth solving? Leave a note — it comes straight to us.

Have a targeted problem?

We'd like to hear about it — a direct, jargon-proof conversation, no pitch.