How We Work: We work in pods - small, dedicated teams embedded with a single enterprise customer, building their voice agents from first call to self-sufficiency. A pod owns the full lifecycle: agent shadowing, conversation design, build, production go-live, and the handoff that makes the customer self-sufficient so the pod can move on to the next one. Each pod pairs forward-deployed engineers with a product manager who owns both the customer outcome and the product learning that comes from it. Every PM on this team runs at least one pod directly and has visibility across others. You're in the customer's world - on their calls, in their systems, solving their problems - while simultaneously shaping what the product becomes. The learnings from your pods feed directly into the platform roadmap and into repeatable, productized assets that scale beyond any single customer. What This Role Owns: You run one pod directly - one of our most complex customer engagements - and float across two others, providing product direction and unblocking teams. But the pod work is only half the job. Your primary technical focus is the conversational experience itself: dialog quality, turn-taking, interruption handling, barge-in behavior, and endpointing. You collaborate directly with the platform voice team on voice pipeline innovation, the low-latency infrastructure that makes a voice agent feel like a conversation rather than a bot reading a script. When the platform team is making stack decisions about real-time streaming, LLM orchestration, or voice-to-voice model integration, you're in the room with production customer data and real-world edge cases that no one else has. You also own an industry vertical. As we scale across customers, you evaluate which patterns, intents, and conversation flows can be productized into out-of-the-box voice content for a specific industry. You're turning one-off wins into repeatable assets, and you have the customer exposure to know which abstractions actually hold. Why This and Not Somewhere Else: You're not choosing between startup speed and enterprise scale. You get both. This team ships voice AI in production with real enterprise customers, with the same urgency and ownership you'd find at a startup, backed by a platform that supports 85% of the Fortune 500. You have a direct seat at the platform table, influencing voice infrastructure decisions with data from your own deployments. This role puts you in a customer's contact center debugging a live voice agent one day and shaping platform architecture decisions the next. That combination is rare, and it's why this team exists.
What You Bring: 12+ years of product management experience, with recent hands-on work shipping voice AI or conversational AI products. What you've shipped in the last 12-18 months matters more than total years. Fluency with the voice stack: ASR, TTS, streaming pipelines, latency optimization, endpointing, barge-in, turn-taking. You've made real tradeoff decisions here, not just evaluated vendor demos. Experience with telephony integration: SIP, call transfers, IVR systems, or CCaaS platforms. Experience working directly with enterprise customers on technical deployments - not from a distance, but in their environment, on their calls, solving their problems. Strong opinions on what makes a voice conversation feel human, backed by experience building or evaluating conversational systems. Technical depth in LLM orchestration for real-time conversational use cases. Even Better: A track record of productizing customer-specific solutions into repeatable, industry-level assets. Industry expertise in verticals with high contact center volume: telecommunications, financial services, travel, healthcare. Experience defining voice-specific evaluations such as task completion, latency, transcription accuracy, and conversational quality.
For positions in this location, we offer a base pay of $190,900 - $334,100 , plus equity (when applicable), vari
Every tech & IT company hiring across India — with AI match scores — on one live map.
Open the map →