📰 Key Takeaways

Klaviyo has agreed to acquire Agency, a three-year-old AI customer service startup founded by serial entrepreneur Elias Torres, on undisclosed terms. Founded in 2023, Agency had raised $32 million from investors including Sequoia, Menlo Ventures, and Felicis before the acquisition.

Under the terms of the deal, Torres will join Klaviyo as Chief Product Officer, bringing along Agency’s 25-person team to accelerate development and expansion of Klaviyo’s two AI agent products: Composer, which builds marketing campaigns, and Customer Agent, which handles post-purchase support like returns and order tracking. Klaviyo co-founder and CEO Andrew Bialecki says the goal is to roll out the combined product to Klaviyo’s 200,000 business customers, with hopes of scaling to millions in the coming years.

For Torres, joining Klaviyo has deep roots. He co-founded Performable (acquired by HubSpot in 2011), then spent eight years as CTO at Drift until it sold to Vista Equity for $1.2 billion in 2021. Back in 2010 at Performable, Torres hired Bialecki — just two years out of Harvard — as one of the company’s earliest engineers, and mentored him through his early startup career. After leaving Performable, Bialecki bootstrapped Klaviyo on his own, and when he raised outside funding for the first time in 2015, he brought Torres in as an angel investor in the seed round. Klaviyo went public in September 2023 at a $9.2 billion valuation. Even with Klaviyo’s stock weighed down by the broader SaaS slump, both men believe the years of customer data the company has amassed will give its AI agents an edge against competitors like Decagon and Sierra.

💬 JudyAI Lab Take

From where JudyAI Lab sits, what’s notable about Klaviyo’s acquisition of Agency isn’t the price tag — it’s that founder Elias Torres was literally Klaviyo co-founder Andrew Bialecki’s career mentor, a relationship going back to their Performable days in 2010.

This points to something bigger: the AI customer service race is shifting into an “acquire for talent and customer data” phase, rather than a pure model-capability contest. Klaviyo isn’t just buying Agency’s 25-person team — it’s feeding years of accumulated marketing data into its AI agents (Composer for campaigns, Customer Agent for post-purchase support) to differentiate against AI customer service startups like Decagon and Sierra. For AI builders, the lesson is clear: as underlying model capability flattens out, the real moat is shifting toward “who owns the proprietary data you can feed the AI” — and product-side data assets are becoming more valuable than the model itself.

If you’re building an AI agent product, it’s worth asking: can the user data you’ve accumulated become a training advantage nobody else can replicate?

📅 Original Article Info

🔗 Further Reading