Jeremy Sullivan

About Jeremy Sullivan

I build AI-powered systems that turn unstructured business data into structured, actionable output.That thread runs through everything I\'ve done over the past 15 years - from delivering the first AI service to production at a public company, to co-founding a product that transforms raw audio into business intelligence, to architecting an AI-driven brokerage platform as CTO.The pattern is consistent: find the business process where humans are manually processing unstructured data, design an AI pipeline to extract structure and meaning, and put human judgment at the decision points while automating everything around them.At Okta, that meant replacing a $100K/year vendor with an LLM-powered lead routing service that outperformed it, building AI-powered compliance tooling for SEC auditors, and winning a company-wide security prize for embedding-based vulnerability detection. At Portana, it meant building Echo Insights - a platform that takes a conversation recording and produces six parallel AI analyses without any technical expertise from the user. At Pioneer Square Labs, it meant decomposing an entire insurance placement workflow into bounded contexts and designing AI automation for each stage.I care about building things that actually run in production - not demos. I think the most interesting work in AI right now is happening at the intersection of frontier language models, domain-driven design, and real business operations. That\'s where I work.

Experience

1. Ai Engineering Leader

   Pioneer Square Labs

   Nov 2025 — Present · Seattle, WA, US

   Lead engineering strategy and hands-on development for a stealth company, on an AI first platform built to drive optimizations and efficiencies.Decomposed the full customer placement workflow - from client intake through close - into six bounded contexts with approximately 30 capabilities. Made build/buy decisions for each, layered AI automation where it created the most leverage, and designed human checkpoints at every decision point.Built an automated lead scoring pipeline that assesses new clients the moment they submit an intake, combining rule-based asset analysis with Claude AI to score, prioritize, and deliver actionable team notifications - surfacing the highest-value opportunities so sales can act first.Designed AI-powered document extraction that converts unstructured client documents into structured entity data with conflict resolution, fuzzy matching, and deduplication - reducing 60 minutes of manual data entry to an automated review-and-approve flow.

Education

• Central Washington University

  Master of Science (MS)

  1996 — 1998

Skills

• Information Architecture
• Mobile Applications
• Microsoft Excel
• Ajax
• Databases
• Docker
• Ec2
• Microsoft Word
• Actionscript
• Microsoft Office
• Ubuntu
• Mobile Devices
• Web Design
• Css
• Mysql
• Cloud Computing
• S3
• Javascript
• Html 5
• Solr
• Jquery
• Web Applications
• Wordpress
• Drupal
• Microsoft Sql Server
• Sql
• Microsoft Project
• User Experience
• Web Development
• Html
• Php