Michael 'Yaro' Yaroshefsky is the VP of AI and Chief AI Officer at Usercentrics, where he leads AI strategy, product development, and the continued buildout of MCP Manager, the MCP governance platform he founded. He s a contributor to the Model Context Protocol specification and writes about AI infrastructure, privacy, and enterprise AI deployment. Before that, he spent a decade as a founder, and before that, he led $273M in software investments at Insight Partners.
The people who are actually getting their hands on open-source models are seeing the future more clearly than everyone else. Most enterprise AI work doesn't require state-of-the-art intelligence. And the economics are about to flip.
87% of Americans were asked to submit biometrics this year. 63% had serious concerns. The people most fluent in technology are the most skeptical of handing over their permanent biological identifiers to companies with imperfect security track records.
Somewhere in mid-2024, our revenue acceleration slowed despite shipping customer-requested capabilities at a lightning-fast rate. It wasn t sudden; it was like a gradual, invisible friction stole our energy slowly over a few months. Confounding matters, users were spending even more time in the platform and 1-day retention figures were noticeably higher. But users were converting [ ]
If you’ve ever been told that sheer willpower is the magic ingredient for success, you ve been fed a lie. Contrary to the popular mantra, willpower isn’t the secret ingredient—it’s the environment that does the heavy lifting. In fact, research consistently shows that the way you set up your surroundings can be far more influential than [ ]
Imagine a world where your AI assistant does more than just generate content—it takes initiative, plans, and executes tasks on your behalf. This is the promise of agentic AI, a groundbreaking evolution in artificial intelligence that aims to free us from mundane tasks and let us focus on what truly matters. Yet, despite its incredible [ ]
Hot take: Don’t be data-driven. Be data-informed. It is pedantic, and what I’m about to say may seem like common sense, but the subtle framing makes a difference. Saying you are “data-driven” suggests that the data makes the decisions for you. It sets the expectation you’ll go where the data tells you to go. It [ ]
There’s something so motivating about spending time directly meeting customers who live by a product you made. It’s more than enough compensation for the challenges that go into building a product: the hard decisions, late-night fire drills, and questions to which there are no easy answers. Leading a company is rarely easy. Elon Musk referred [ ]
I recently came across an interesting debate about the future of the Scrum Master role in the tech industry. In fact, Alex Ortiz and I just discussed this today while recording a new episode of Summer of Atlassian. There seems to be a feeling that this role is fading away, with its responsibilities increasingly being [ ]
Has the lean startup methodology circa 2011 outlived its prime? I think it is overdue for an update. If you’ve heard the term “minimum viable product” or MVP, you’re already familiar with one of the key concepts. Originally published by Eric Ries in 2011, this methodology suggests that startups should find market fit by building [ ]
It’s Friday afternoon so I’ll keep this light— I just read a WSJ article about pets involved in the hiring process. I can’t promise that you will get hired for one of Visor s open roles if you send me a picture of your cat, but I *can* promise it will increase the odds. 🐈 Anyway, [ ]