Apollo.io Innovation & Technology Culture

How does your team stay ahead of emerging technology trends while scaling fast?

It’s easy to be caught up on hype. The best method for us is to focus on the vertical use cases we want to empower with AI — e.g. find prospects or send an email — and pick the most efficient and high-quality implementation regardless of the new technology. For example, we ran blind testing across models for our AI messaging use case and found Claude Haiku 4.5 outperformed larger, more expensive models for our cold outbound email use case.

What recent product or feature are you most proud of — and what impact has it had?

Our AI Assistant. We rebuilt it from a 20-node system to a React-style agent architecture, analyzed 1,750 real user conversations to fix failures, and now have 40,000 weekly active users driving 2.3 times more meetings in their first 14 days. As one of our users, Delaney Beam, put it, “I’m no go-to-market engineer, but Apollo’s AI Assistant makes me look like one.”

How Phillips Creates a Culture Rooted in Innovation and Experimentation at Apollo

• We let our users lead us to product-market fit. When we launched AI Research, which is now at 55 million research actions per month, customers built use cases we never imagined. One found website translation gaps, and another analyzed competitor strategies. The fastest way to achieve product-market fit is replicating what our users are already doing with better UX.
• We hire for domain expertise and agency over anything else. Our product builders on the team started as AI annotators who submitted more than 100 product bugs before joining the AI team full time.
• We build feedback loops that never close. We review conversations on a weekly basis — up to 1,750 — use the product daily, and learn from real user scenarios to know what we should build next.