Built
Products and open source
The products I build, then the open-source tools and early concepts behind them. Some are shipped and production-tested; others are ideas I am working through with a few engineers and domain experts.
Products
Company I co-founded
Personize
Know each customer. Make it personal.
Personize turns unified customer memory into contact-level experiences across pages, emails, print and sales workflows, with governance and private deployment options.
- Governed customer memory as the base layer
- Contact-level experiences, not segments
- Private deployment when data has to stay in your environment
Personize product
GPE, the Generative Personalization Engine
One story. Every channel.
GPE carries what a team knows about a person into the experiences it creates for them: the page, the email, the sales conversation and beyond.
- Bring the context together
- Shape the next experience and put it into the work
- Use approved knowledge, keep review where it belongs, make delivery observable
Personize product
Personize Private
Your memory. Your environment.
A unified customer memory layer for teams that need control over where it runs, how it is governed, and which systems can use it, through private MCP and API interfaces.
- Runs in a customer-owned environment
- One context layer for every agent and system
- Memory Box: the same memory on local hardware, discussed case by case
Personal projects
Personal project
Peers
Tell your AI who you want to meet.
Meet people through ChatGPT or Claude, with no new app to learn. Your AI and the Peers engine find matches across communities, and Peers emails both people an introduction when the fit is mutual.
- Works inside the assistant you already use
- Matches on shared goals and complementary skills
- You control the pace of introductions and your privacy
Personal project · Early Access
Google Docs for Agents and Humans
For AI agents and the people working with them
A shared document layer for AI work: agents publish what they know, humans comment, other agents read it, and the same live link keeps evolving.
Open source and tools
- CRM AI OperatorsFor: Revenue teams, and the AI agents working for themOpen Source
Point a capable agent at your CRM and it works at ten records and falls apart at ten thousand. CRM AI Operators is the pattern that fixes that: a subagent per record, grounded in what the org knows and the rules it works under.
- Generative SitesFor: CMO, Demand Gen, Performance Marketing, Web TeamsEarly Access
Drop a script tag, mark elements, get AI-personalized website copy per visitor. Property lookups, structured generation, and visitor deanonymization from a single <script>. Works on WordPress, Webflow, AI-built sites, custom code, and authenticated SaaS. The idea: every page becomes a 1:1 conversation grounded in what you already know about the visitor, account, or customer segment.
- SignalFor: Product Manager, CX Leader, Lifecycle MarketingActive
A notification engine that decides IF, WHAT, WHEN, and HOW to notify across in-app, email, SMS, and product nudges. Scores every event against entity memory and governance rules, then sends, defers, or skips. Self-improving: captures open rates, click-throughs, and reactions to tune its own scoring over time. The goal is reducing notification volume by 40-70% while increasing engagement on what actually sends.
- Data ExtractionFor: Operations Leaders, Data Engineering, RevOps, Companies with massive inbound databasesActive
Paragraph-by-paragraph AI extraction across documents, transcripts, emails, and reports at scale. Maps unstructured text into structured properties (objections, buying signals, risk indicators, renewal concerns, competitive mentions, action items) and entity-tied memories. Extract once, store as durable memory, reuse across every downstream workflow rather than re-paying token cost on every query that touches the same source.
Exploring
Early concepts and builds, tested with a small group of engineers and experts before they harden into products.
- Record WorkspaceFor: AI Engineers, Account Managers, CSM Leaders, Deal TeamsExploring
An AI-native CRM where the primary user is an agent, not a human. Each record carries its own workspace: notes, tasks, agent messages, status, memory links, workflow history. Multiple agents (research, qualify, write outreach, handle replies) operate on the same record without losing context or stepping on each other. The bet: agents do not only need tools, they need a place to work.
- Memorize[Examples]For: AI Engineers, Agent Builders, Applied AI Teams, ML/AI LeadsExploring
A memory of approved and rejected agent outputs paired with the input context, the reasoning, and why a human accepted or rejected the work. New tasks retrieve the closest real precedent instead of improvising from a synthetic few-shot. Surfaced to agents over MCP. Useful where the right answer is subjective: writing a sales email, deciding to send a notification, classifying lead quality, handling a sensitive reply.