RESEARCH & ML•Bengaluru (Indiranagar)•Full-Time
Lead Audio ML & Speech-to-Speech Researcher
Pre-train and distill foundation speech models for sub-150ms full-duplex conversational audio.
Speech MLPyTorchAudio Tokenization
4+ yrs MLApply for Role
▶CAREERS
Building human-grade voice AI.
We're a small, high-density team of machine learning scientists, telephony engineers, and real-time systems architects building full-duplex conversational models for millions of phone calls.
CORE TEAM • BUILDING HCM-1BENGALURU & GURUGRAM
CRAFT & SHIPSUB-180MS SPEECH
CORE TELEPHONY & SIP•Gurugram (Cyber City)•Full-Time
Principal Telephony & SIP Gateway Engineer
Architect high-throughput SIP carrier gateways and WebRTC audio proxies handling 50k+ concurrent calls.
RustSIP / RTPWebRTC
5+ yrs TelephonyApply for Role
REALTIME PIPELINES•Bengaluru (Koramangala)•Full-Time
Senior Real-Time Pipeline Engineer
Build low-latency audio streaming pipelines between carrier SIP trunks and inference clusters.
PythonGoWebSockets
3+ yrs BackendApply for Role
FULL-STACK & DX•Bengaluru (Indiranagar)•Full-Time
Staff Full-Stack & Developer Experience Engineer
Develop our Prompt Lab IDE, audio waveform interfaces, and public client SDKs.
TypeScriptNext.jsSDKs
4+ yrs Full-StackApply for Role
SOLUTIONS ENG•Gurugram (Cyber City)•Full-Time
Enterprise Solutions Architect & Telecom Lead
Deploy voice agents, PBX trunk bridges, and CRM integrations for high-volume enterprise clients.
SIP TrunksEnterprise PBXCRMs
3+ yrs SolutionsApply for Role
RESEARCH & ML•Bengaluru (Indiranagar)•Full-Time
ML Inference & CUDA Kernel Optimization Engineer
Write custom Triton and CUDA kernels for streaming continuous speech generation on GPU clusters.
CUDATritonTensorRT-LLM
3+ yrs CUDA / SystemsApply for Role
GENERAL APPLICATION
Don’t see your exact role?
We are always looking for exceptional audio ML researchers, low-level Rust/C++ kernel hackers, and SIP telephony experts. Send your GitHub, code samples, or past project writeups to our founding team.