Before I graduated from college in 2007, I bought a failing restaurant.
It was the Pita Pit near the University of Washington campus. I was 22, and for the next couple of years I worked fourteen-hour days turning it around. At 24 I bought a second location in Bellingham. I sold both, moved to Hollywood, and started a med spa. That one failed.
Failing is usually where the story gets interesting, and mine did. That chapter led me to the woman who has now been my wife for fourteen years, and it led me to a small company called RingLead. It was the first W-2 job I ever had.
For the fifteen years since, my title has said sales. I have been an individual contributor, a sales leader, and a chief revenue officer. I never stopped being an owner.
While RingLead was still early, I invested in a seafood company doing $30 million a year in international sales. During a stretch when RingLead went through several CEOs, I spent a couple of years working there as a 1099 contractor, effectively running my own book inside someone else's company. Along the way I started a property company and an e-commerce business on the side. I have never been able to take the entrepreneur out of me, and I stopped trying a long time ago.
What the sales title did give me was a front-row seat to one problem, over and over, for fifteen years.
Companies lose revenue through records nobody can keep true.
It never happens dramatically. Nobody gets fired over a stale title field. The champion who left eight months ago is still the primary contact on a renewal. Two reps are working the same account under two spellings of its name. A routing rule stops firing because an integration wrote USA into a field that expects United States, and nobody notices until a quarter's worth of inbound has gone to the wrong queue. Each one is small. Added up across every record in every CRM, they are the difference between a revenue team that trusts its system and one that quietly works around it.
At RingLead I was part of taking the company from early stage to its acquisition by ZoomInfo in 2021, and as an individual contributor I closed roughly thirty percent of the company's revenue by the time we sold. What that really gave me was thousands of conversations with the people who own CRM data: operations leaders, marketing ops teams, admins who had inherited systems nobody could fully explain anymore. I heard the same problem described a thousand different ways, and I carried it back into the product. Despite the sales title, I was heavily involved in RingLead's product design and direction, which is where I learned that the distance between what a customer describes and what software actually does is where most products are won or lost.
After the acquisition I ran ZoomInfo's data-as-a-service business, which grew from $40 million to $160 million in 27 months. That was the other side of the same problem. We were selling external truth: who works where, what a company looks like today. And I watched customers buy that truth and then struggle to make it last inside their CRM. The data was right the day it arrived. The records it landed in started drifting the day after. I spent a lot of my time at ZoomInfo with the product team, and that gap was never far from the conversation.
I did not start LiveCRM because an idea arrived one morning. I have carried this conviction for most of my career. What changed is that two things showed up at the same moment.
The first is the technology. AI can now reason about a record in context, and tools like Claude Code have changed what one person can build. For fifteen years I shaped products from the sales side of the table, handing what I learned to the people who would build it. Now I design the architecture of LiveCRM and I write the code. In four months it has gone from nothing to a deployed platform running against live Salesforce and HubSpot.
The second is the buyer. Every revenue organization is putting AI on top of its CRM, and an agent does not question the record it is handed. It acts on it, confidently and at volume. Bad data used to slow people down. Now it gets amplified. Leaders who once treated data quality as housekeeping are discovering that their AI plans depend on it, and they finally have a reason to fund fixing it properly.
LiveCRM keeps a live view of what your CRM believes, compares it against what is true in the outside world, and turns the gaps into action: deduplication, normalization, enrichment, routing, and the workflows that should follow when something changes, like a key contact leaving an account. Every change runs under rules you set, proposed for your approval or executed within limits you define, and every one of them is audited and reversible. It works with Salesforce and HubSpot, and it exposes the same capabilities to AI agents, so the agents your team is deploying can act on the CRM through a governed path instead of around one.
Almost twenty years ago I was working fourteen-hour days in a pita shop because I wanted to own something. LiveCRM is the same instinct, pointed at the problem I know better than anyone.
I will write here about what I am learning as I build: how CRMs actually drift, what it takes to let software change a system of record safely, and what I see from the people who run revenue operations every day. Some of it will be technical. All of it will be honest about what works and what doesn't yet.
If your CRM is about to carry more weight than it was built for, I would like to hear about it.
If this sounds familiar
The readiness scan reads your records and reports what it finds. It writes nothing, and you keep the report.