AI Calling

The problem

For many businesses the phone is still where customers ask, complain and buy, and a large share of those calls are routine: the same questions, follow-ups and reminders, handled one at a time. That takes hours small teams do not have, and it is hard to scale without hiring. AskTalos was building an AI communication platform, and voice calling was the channel where automation could take the most repetitive work off people's plates.

How I approached it

As the sole product manager I owned the roadmap and ran the full cycle from user discovery to launch. I began with conversations with the people who make and receive these calls, listening for the situations that repeat most and the moments where a human is genuinely needed. Those conversations decided what the first version should handle automatically and what it should leave alone.

I defined scope and priorities with engineering, design and business stakeholders and made trade-offs openly to keep shipping on tight timelines. The module launched as part of a wider suite, so I made sure calling shared customer data with the CRM and the messaging channels instead of living on its own.

The outcome

AI voice calling launched, letting businesses handle customer calls automatically, without human agents, and taking routine communication off their teams.

What it taught me

Automation earns trust when it knows its limits. Deciding what the AI should not do mattered as much as what it should.