// product · data · engineering
Hi, I'm Mike
Senior Product Manager | Enterprise Platform
20+ years building data-driven products — from analytics pipelines and machine learning systems to enterprise SaaS platforms used by millions. I bridge product strategy with hands-on software engineering and empowering leadership.
// who i am
About Me
With over 20 years across product management, data analytics, and software engineering, I've led platform and analytics initiatives at Semrush, Sky, Snowplow, Ometria, and Webtrends — scaling products used by millions of users and building data systems from the ground up.
I stay hands-on: writing Python and SQL, building ML models, and shipping side projects for fun (see below). That mix of strategic product thinking and real technical depth is what I bring to every team I join.
// live demo
Ask My Agent Anything
Have a role or project you want to check my fit for? Ask below.
// selected work
Career Highlights & Projects
A mix of the hard problems I've led teams through at work, and the personal projects I build to stay sharp and curious.
Current Role Highlights
Report Studio
Challenge: Our legacy reporting tool was built for speed, not scale — no cross-product data, filters that reset, elements that quietly went stale. Fixing it meant moving hundreds of live enterprise clients off years of built-up reports.
Approach: Led the cross-product rebuild and designed an honest, opt-in migration path. When manual rebuilds proved too slow at scale, I built tooling to auto-match legacy elements — pushing real coverage from a flawed 54% estimate to a verified 87%.
Impact: 50-client beta to GA with over 50% adoption in the first month.
"Good migrations aren't about hiding the hard parts — they're about giving people real tools to move forward."
Element-Level API
Challenge: An internal-only API meant to power our own dashboards had been quietly adopted by power-user enterprise clients pulling raw data into their own BI tools — no rate limits, no docs, and a handful of accounts driving most of the load.
Approach: Built a usage-limits strategy from real traffic and cost data, then aligned engineering and commercial teams around a phased rollout — warnings before enforcement, with a clear upgrade path instead of a cutoff.
Impact: Protected $500k+ in enterprise MRR and avoided ~$10k/month in infrastructure cost, with limits signed off up through the CPO.
"The best fix for an accidental product isn't to kill it — it's to give it real guardrails and a real home."
Enterprise Collaboration Suite
Problem: SEO teams needed deep collaboration between internal teams, the platform, and external experts — not a bolt-on help widget.
Solution: Led development of an integrated suite combining chat, Kanban tasks, meeting booking, smart docs, and an expert marketplace, navigating build-vs-buy across 6+ tool categories.
Impact: Renewing clients used the suite more than clients who churned, and it was one of the highest-rated tools with our support team — later sunset in favor of Intercom as the business consolidated support tooling.
"Proof a fast-moving team can build something clients genuinely value, even when the business later chooses a different path."
In the Media
Conversations and writing on product strategy and AI, elsewhere on the internet.
Side Projects & Experiments
With AI-assisted tooling, I can test an idea in a weekend instead of a quarter. Not everything here is meant to become a business — most of it is just me staying hands-on with code and following my curiosity. Two get the deeper write-up below; a few smaller ones are linked at the end.
Predictive ML Experiments
Challenge: Building full end-to-end data pipelines in notoriously difficult prediction domains — horse racing and forex trading.
Applications: A horse racing predictor (48% hit rate) that collects race data and trains daily models, plus EUR/USD trading signals (50% profitable entry accuracy).
Learning Focus: Real-world data collection and parsing, feature engineering for time-series predictions, and model performance in high-noise environments.
"Sharpens my data pipeline skillset and satisfies my curiosity in what's possible with data and compute."
Helperee.com
Evolution: Started as a learning project to build a PDF chat tool; now a long-running experiment in agentic project management and data cataloging.
Status: Early and unhurried. I work on it in the background as a genuine long-term bet, not a product I'm pushing to market — it may take years to find real shape, if it ever does.
Why it's here: It's where I test AI agent patterns with real code rather than just theory.
"A slow-cooking experiment — valuable for what it teaches, whether or not it ever becomes a product."
// say hello
Get In Touch
Hiring for a product management role, have a question, or want to discuss an advisory project? Feel free to reach out to me using the form below, schedule an intro call or connect on LinkedIn.