Senior Python Developer Voice AI, Telephony & Real-Time Systems
About the Role :
We are building an AI receptionist and ambient assistant platform that answers and routes calls, books appointments, summarizes conversations, and provides real-time help inside calls. Youll design and ship backend services and real-time pipelines that power appointment scheduling, voice AI (ASR / TTS / LLM), transcripts & summaries, ambient assistance, vertical solutions (medical, property management, legal), third-party integrations (e.g., Toast, Jobber), and AI-assisted callflows.
What Youll Do :
Real-time voice / telephony services :
- Build Python microservices (FastAPI) for call control, presence, and routing.
- Implement SIP / WebRTC / RTP signaling & media handling, conference control, and callflow orchestration.
- Create low-latency streaming ASR pipelines and TTS playback; manage VAD, diarization, barge-in.
AI assistants & conversation intelligence :
Integrate ASR / TTS / STT vendors (Whisper / Deepgram / Google / OpenAI, etc.) and LLMs.Ship summarization, action-item extraction, and RAG over knowledge bases / Content.Implement prompt safety, guardrails, evaluations, and fallbacks.Appointment scheduling & personal assistant features :
Build booking flows with Google / Microsoft calendars, availability, buffers, reminders, time zones.Capture / relay messages, human escalation, policy-based routing and SLAs.Ambient assistance in live calls :
Join / leave wake phrases; real-time IM / notes; add / remove participants; read knowledge on demand.Vertical & partner integrations :
Adapt flows for healthcare (HIPAA-aware intake, reminders), property management (tenant issues), and legal (intake & scheduling).Integrate with third parties such as Toast and Jobber via APIs / webhooks.Platform engineering :
Design resilient, observable systems (OpenTelemetry), queues / streams (Celery / Redis / Kafka), and data stores (Postgres / MySQL / Redis / Elastic / Vector DB).Own CI / CD, test coverage (pytest), load / soak testing, and production readiness.Collaborate with Product, QA, and Ops for staged rollouts and SLO- driven operations (latency, accuracy, uptime).Must-Have Qualifications :
3 to 10 years professional Python experience building production backends / microservices (FastAPI / Flask).Real-time systems : streaming I / O (websockets / gRPC), asyncio concurrency, low-latency design.Voice AI : integrated at least one ASR and TTS provider and orchestrated LLM calls; built summaries / transcripts at scale.Data & infra : SQL schema design (MySQL), caching (Redis), queues (Celery / Kafka), containerization (Docker), cloud deployment (AWS / GCP / Azure).Testing & observability : pytest, contract tests, performance profiling, Security & compliance mindset : secrets management, PII handling, auditability; awareness of HIPAA / PCI / SOC2 controls.Nice-to-Have :
Telephony / RTC : hands-on with SIP / WebRTC / RTP, media servers / gateways conferencing, DTMF, transfers.LLM ops : prompt / version management, eval harnesses, cost / latency tuning, vector DBs NLU / NLP pipelines (spaCy), structured outputs with Pydantic, Calendar & booking APIs (Google / Microsoft), CRM / Helpdesk (Salesforce, HubSpot, Zendesk).Event-driven design (outbox, idempotency, saga) and blue / green or canary rollouts.Frontend familiarity (TypeScript / React) to collaborate on operator consoles.Tech Stack :
Python 3.11+, FastAPI, Pydantic, asyncio; WebSockets / gRPC; SIP / WebRTC / RTP; Asterisk / FreeSWITCH / Janus / OpenSIPS; Redis, Postgres / MySQL, Kafka; Docker, Kubernetes;, Terraform; OpenTelemetry, Grafana; pytest.Soft Skills :
Bias to action, clear written and verbal communication, production pragmatism.Partner-friendly mindset for vertical integrations.Cross-functional collaboration with Ops / QA / Product and customer-facing teams.(ref : hirist.tech)