Jayesh Parmar
Software Engineer at LiveKit
Ahmedabad, Gujarat, India
Summary
🤩
Rockstar🎓
Top SchoolJayesh Parmar is a software engineer with five years of experience building scalable, concurrent systems for healthcare and real-time multimodal AI applications. He shipped high-impact features at HHAeXchange—like an IVR system handling millions of monthly calls and a Twilio migration—and now works at LiveKit focusing on real-time voice agents and STT integrations. Proficient in AWS serverless, C#, .NET, TypeScript and backend integrations, he contributes to open-source projects such as agno and LiveKit where he implemented Twilio/Cal.com tools and improved function-calling and VAD for agents. His strength is turning regulated, high-throughput requirements into reliable services while bridging integrations between telephony, calendar, and AI tooling. A Dharmsinh Desai University B.Tech in Computer Engineering graduate, he actively seeks challenges that combine healthcare impact with cutting-edge AI and realtime systems.
5 years of coding experience
2 years of employment as a software developer
Bachelor of Technology - BTech, Computer Engineering, Bachelor of Technology - BTech, Computer Engineering at Dharmsinh Desai University
Github Skills (19)
application-framework10
apim10
realtime10
python10
twilio-api10
app-framework10
llm10
agent10
api10
openai10
ai10
web-framework10
http-request9
google-cloud9
developer-tools9
Programming languages (2)
Github contributions (5)
livekit/agents
Oct 2024 - Apr 2025
A powerful framework for building realtime voice AI agents 🤖🎙️📹
Role in this project:
Back-end & ML Engineer
Back-end & ML EngineerContributions:172 reviews, 163 PRs, 700 pushes in 6 months
Contributions summary:Jayesh made significant contributions to the `livekit/agents` repository, focused on integrating support for various speech-to-text (STT) models, including Groq, Google, and FAL's Wizper. They implemented and refined functionalities related to function calling within the agent framework, addressing issues, such as exception handling and correct parsing of function argument types. Additionally, the user added metrics and functionality related to voice activity detection (VAD), improving the agent's real-time audio processing capabilities.
agentsaiopenaireal-timevideo
agno-agi/agno
May 2024 - Nov 2024
A lightweight library for building Multimodal Agents. Give LLMs superpowers like memory, knowledge, tools and reasoning.
Role in this project:Back-end Developer & Integration Engineer
Back-end Developer & Integration EngineerContributions:11 reviews, 5 PRs, 21 comments in 6 months
Contributions summary:Jayesh primarily contributed to integrating the Twilio and Cal.com APIs as tools within the "agno" framework. They developed a `twilio.py` file with classes and methods for sending SMS messages, getting call details, and listing messages, as well as a `calcom.py` file to manage calendar events. The user also created example usage in the `cookbook` directory. The user's commits demonstrate a focus on tool development, API integration, and potentially building automation workflows for the agent.
python-librarypythondatadata-sciencellmops
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.