AI product leadership, from prototype to org-wide adoption.
I'm Ryan Smith, a VP of Product in data-heavy platforms. I prototype the hard part myself to de-risk the bet — then build and lead the team that takes it to production and adoption across the business. Eleven years in product — today leading a nine-product-line data & AI platform portfolio.
Ryan Smith — VP Product, Kinesso (Omnicom). Leads an 8-person product org across 4 countries and owns a 9-product-line data & AI platform. Built the AI product strategy ground-up: an org-wide AI assistant (3,500+ users) and a production text-to-SQL platform live across 60+ markets, supporting millions in revenue and retention.
Numbers I can stand behind.
Markets where the NL-to-SQL platform runs in production — supporting millions in revenue and client retention
Active users on the AI assistant I took from prototype to org-wide adoption
Brief-to-activation in an executive-sponsored agentic-buying pilot
Reduction in data-error rate across global markets after the billing redesign
Daily reporting-and-billing runtime, at the same data volume
Product org built and led across four countries
Four products. Built, shipped, adopted.
A closer look at how I operate — de-risking the hard part myself, then leading the team that's accountable for the outcome.
Agentic buying, from months to days
An executive-sponsored bet to compress the media brief-to-activation cycle — I prototyped the core to de-risk it, then led the team that proved it in a multi-client pilot.
One AI assistant, adopted org-wide
Unified fragmented, siloed AI experiments into a single assistant that follows users across the platform.
Plain-language questions, trustworthy answers
A natural-language-to-SQL agent so non-technical users can query data directly — with transparency and auditability built in.
Six hours to one, at global scale
A reporting-and-billing backend redesign that turned a six-hour bottleneck into one hour and cut the data-error rate roughly 4× across every market.
Don't take my word for it.
I built a live AI assistant and a natural-language-to-SQL demo right on this site — grounded in my real work, with the guardrails shown. Ask it anything, or paste a job description for an honest fit read.
A leader who de-risks by building.
Four principles that hold across every team and product I've owned.
De-risk first, then scale.
I'll prototype the riskiest part myself to prove the bet is real — then build and lead the team that productionizes and scales it. Every agent in my portfolio started this way.
Outcomes you can audit.
Every AI product I ship has transparency, auditability, and a measurable success metric built in from day one — not bolted on later.
Teams that scale with the platform.
I grew my product org from two to eight across India, Malaysia, Argentina, and the US — including two managers who now report to me, spanning nine product lines — coaching from first-time PM to mid-career.
Steady through change.
I led product and people through a major acquisition and integration without losing delivery momentum or the team.
A tool I built, shipped, and then shut down.
I built an AI tool for my own product team that auto-structured incoming Jira tickets — it cut requirement misunderstandings — measured as engineering complaints about ticket quality — by ~27%, a number I was proud of. Nine months in, I killed it.
The bet was simple: normalize vague tickets at intake to kill downstream rework. The metric was real. But I watched some PMs stop verifying the model's output and start trusting it blind — and about a quarter fewer misunderstandings isn't worth teaching a team to outsource its judgment to a black box.
So I shut it down — even though two PMs leaned on it daily — and I changed how I build. Every AI product I've shipped since — including the conversational-analytics platform — shows its work, so trust is earned, not assumed. It's why the live demo streams its steps and refuses what it can't ground. The full story, including what I got wrong.
From the flight line to the platform.
Eleven years in product, almost all of it in data-heavy, high-transaction platforms — billing, reporting, and analytics at enterprise scale. Today I'm VP of Product at Kinesso (Omnicom), where I own a data and AI platform portfolio — nine product lines serving thousands of users, a team across four countries, and two managers reporting to me — and built the AI product strategy from scratch. The portfolio spans roughly 2.4 trillion rows of data.
Before product, I founded and ran a B2B inspection business end-to-end — owning sales, proposals, and delivery, and landing enterprise contracts with Chevron and Kiewit. Earlier, I served as a U.S. Marine Corps avionics technician, held a Top Secret clearance, led up to 40 Marines, and earned the Navy & Marine Corps Achievement Medal.
I advise early-stage AI startups pro bono, and sharpen my craft through Pragmatic Institute product training.
One team, several nameplates: Kinesso was built inside IPG Mediabrands, and Omnicom acquired IPG in November 2025 — I've led the same platform and team through the corporate changes without losing delivery momentum or people.
Vouched for by the people who managed me.
Two of my most recent managers, in their own words.
I'm a strong believer that great product managers need to understand the technology deeply. Ryan is a perfect example of that. I had the pleasure of having Ryan on my product leadership team while we were building IPG's global data and technology platform. He led the modernization of our data warehouse and measurement stack from the ground up, bringing together the technical architecture, product thinking, and client needs behind it. While much of the industry was focused on chatbots, Ryan and his team built a text-to-SQL agentic capability that actually worked at enterprise scale, grounded in a proper semantic layer and connected to real business use cases. That work directly supported millions in revenue for some of our biggest clients across 60+ markets. Ryan combines strong technical depth with real product judgment. He is a great leader, a thoughtful team manager, and someone with an incredible work ethic. I'd work with him again in a heartbeat.
Ryan reported to me for five years, and in that time he became the person I'd hand the hardest problems to. He led the product effort to build our Data Platform from the ground up — not inherited, not a rebuild, from scratch — and it's now the backbone of how our team operates. What I'll remember most is that Ryan was the one who actually dragged us into AI. Not in a buzzword way. He figured out where it genuinely fit, wove it into how the team works day to day, and is now leading our AI initiatives. He's the person everyone routes their AI questions to, and that reputation was earned. I've managed a lot of people. Ryan is the one I'd want and need on my team.
Let's build something that ships.
If you're hiring for AI product leadership — or want to compare notes on shipping AI that holds up — email me, or grab time directly on my calendar.
The conversations I'm most interested in: VP or Head of Product roles with a real AI mandate — especially at AI-native, data-infrastructure, or data-heavy platform companies. Remote (US) or Austin.