About

I’m Piyush Gambhir, a Software Engineer II at Spyne, building production voice AI and telephony infrastructure.

I build voice agents, operate the systems around them, and contribute the fixes upstream.

Most of my work has been around production AI systems, data pipelines, and automation. I currently build and operate conversational AI agents for auto dealerships, with a focus on evaluation, response quality, cost, and latency in real operating environments. Shipping is the easy part. Making systems reliable, debuggable, and economical at scale is the real work.

I've designed and maintained end-to-end data ingestion pipelines, among them browser automation for legacy dealership systems without APIs, large-scale scraping, and inventory normalization. I've also built core CRM components, workflow orchestration, and a dealer website builder with an integrated CMS.

I enjoy working where the ambiguity and the constraints meet: bad data, partial integrations, unclear workflows. My approach is to simplify aggressively, design for failure, and optimize for maintainability.

I'm interested in how AI can be applied pragmatically, not as hype but as infrastructure that improves reliability, speed, and decision-making.

Experience

Software Engineer II

Spyne

May 2026 to Present · Gurugram, India

Leading the migration of production voice agents from Vapi to LiveKit Agents, owning architecture, telephony and SIP integration, and rollout, and collaborating directly with LiveKit’s engineering team along the way.
Contributed upstream to the open-source LiveKit Agents SDK: eight merged PRs across the Python (livekit/agents) and Node (livekit/agents-js) frameworks. They expose OpenAI Responses API controls used in production and preserve the original caller ID on Twilio warm transfers.
Took voice-agent infrastructure cost from $0.26 to $0.10 per minute in our deployment, roughly 62%, alongside the LiveKit migration and the provider and pipeline tuning that followed.
Scaling Spyne’s production conversational AI stack to support increasing traffic, concurrency, and reliability requirements, and strengthening it through disaster recovery, failure handling, and production readiness work.

Software Engineer I

Spyne

Jan 2024 to May 2026 · India

Built Spyne’s production conversational AI stack from zero to one on Vapi, powering dealership lead handling, follow-ups, and customer conversations, then began laying the groundwork for the transition to LiveKit.
Improved agent responsiveness and conversation quality through prompt optimization, streaming pipelines, turn-taking improvements, and infrastructure tuning.
Built the core telephony layer across Twilio, Vonage, and Telnyx, including caller identity, CNAM lookup, STIR/SHAKEN support, banned-number filtering, and A2P 10DLC workflows.
Shipped full-stack systems across inventory ingestion, dealer websites and CMS, CRM workflows, lead management, and automated customer follow-ups.
Rebuilt my development workflow around coding agents: spec-first delegation, parallel runs, and verification gates before accepting a diff. Monthly commits across 15 production repositories went from 50 to 90 up to roughly 300.

Open source, informed by production.

My LiveKit Agents contributions expose OpenAI Responses API controls used in our voice pipeline, including service tier, output-token limits, and usage reporting. They also let a Twilio warm transfer show the supervisor the original caller’s number instead of the business line. The merged pull requests are the public record of that work.

Python SDK contributions ↗Node SDK contributions ↗Full résumé →

Experience

Software Engineer II

Software Engineer I

Open source, informed by production.

Merged into LiveKit Agents