Post by Insight Partners

This month, Onsite Hour launched our AI Innovators series, highlighting how ScaleUp leaders across our portfolio are using AI in transformative ways. Last week's session explored one of the defining challenges for software operators right now: how to move beyond isolated AI pilots and build a company-wide capability that actually scales. Stay tuned as the series continues through early June! Led by Ajay Gandhi and Santosh Iyer from Insight Partners, and joined by Nick Laird and Robert Rainey (VenturEd Solutions), Jason Walker (BigPanda), and Collin Rhea and Chad Ludwig (Aptean), the session drew on real operating models from companies at different stages of the AI journey. We covered: • AI innovation at company scale needs to impact product, GTM, and operations simultaneously to drive growth and margin gains. • AI transformation needs a named owner with real authority, someone who coordinates across business units, owns the P&L impact, and breaks bottlenecks. • Use case prioritization is more than impact and risk. Before committing resources, evaluate whether you have the right people, optimized processes, available tools, clean data, and an organization ready to embrace the change. • Scaling AI in one function without lifting the others creates new bottlenecks. The whole system moves at the pace of its slowest part. Finding and targeting that constraint is where the real gains are. • The shift that separates early adopters from mature operators is moving from AI assisting humans to humans orchestrating AI. Teams that get this right focus on defining intent, setting guardrails, and validating output rather than doing the work themselves. • Design your AI systems for the models you will have in six months, not the ones you have today. Building ahead of the capability curve means upgrades make you faster rather than forcing you to rebuild. The core message from the session: Scaling AI is an operating model problem, not a technology problem. The decisions that matter most are about ownership, measurement, and how work actually flows. We've got 4k+ open ops, product, and engineering roles across the Insight portfolio — join teams where AI stops being a pilot and starts being the operating model: https://lnkd.in/gw2yUrGP