Sierra Acquires Takeoff to Build Long-Horizon AI Customer Service Agents

What happened

Sierra, the AI-powered customer service platform, has acquired Takeoff, a startup specialising in long-horizon AI agents — systems capable of executing complex, multi-step tasks autonomously over extended periods rather than responding to single, discrete queries. The deal signals Sierra's intent to move beyond reactive, turn-by-turn customer interactions and towards AI agents that can independently manage prolonged service workflows from start to finish.

Takeoff's technology is designed to enable AI agents to plan, reason and act across sequences of decisions without requiring constant human prompting. By integrating this capability, Sierra is positioning itself to offer enterprise clients AI agents that can handle genuinely complicated customer service scenarios — think multi-day returns investigations, complex account resolutions or iterative troubleshooting — rather than the scripted, single-exchange responses that have characterised most deployed conversational AI to date.

Why it matters

The acquisition reflects a meaningful shift in how AI is being applied to customer experience. Most enterprise AI deployments today remain shallow: they answer FAQs, route tickets or handle simple transactions. Long-horizon agents represent a fundamentally different proposition — one where the AI takes ownership of an outcome rather than merely responding to an input. For service designers, this changes the unit of design from the individual interaction to the entire resolution journey, demanding new thinking about trust, handoffs, escalation logic and customer transparency.

From a behavioural economics perspective, the stakes are high. Customers extend trust incrementally; an AI agent that autonomously manages a multi-step process on their behalf must earn and maintain that trust at every stage. Errors or opacity mid-journey can cause disproportionate damage to brand perception — the psychological cost of a failed autonomous action is considerably higher than a failed single-turn chatbot response. Organisations adopting long-horizon agents will need to invest as heavily in the experience architecture around these agents as in the technology itself.

By the numbers

The available sources do not disclose deal terms, funding figures or headcount data for this acquisition.

The Renascence take

The industry conversation around this acquisition will almost certainly focus on capability — what these agents can now do. That framing misses the harder, more consequential question: what happens to customer trust when something goes wrong mid-journey and no human was visibly in the loop?

Long-horizon AI agents are not simply faster chatbots — they represent a transfer of agency from the customer to the machine, and customers have not yet been asked whether they consent to that transfer. The behavioural principle at stake is perceived control: people tolerate poor outcomes far better when they feel they had a hand in the process. Customer-obsessed operators adopting this technology should design explicit "check-in" moments into autonomous workflows — not as a technical fallback, but as a deliberate trust signal. The goal is not seamlessness at all costs; it is confidence at every stage.