About Winter

Winter started as PeerIslands in 2018, trusted by large enterprises to modernise their most critical systems in banking, healthcare and telecom. Today we’re an AI-native company backed by $64.5M in funding, with around 500 people across the US and India. At Winter, the people who decide what to build are the people who build it, with no handoffs and no decks that go nowhere. Our work goes live in industries where it matters, powered by our own AI platform, The Protean, which runs 100+ agents.

About the Role

Most AI work today never gets past the demo. Ours runs on real phone calls with real customers, where 300ms of latency is noticeable and a missed interruption breaks the conversation.

We’re hiring a Senior AI Engineer to design, build and ship production-grade GenAI applications, specialising in real-time voice agents. You’ll own the full stack: backend services, web dashboards, LLM orchestration, RAG, telephony, speech recognition and synthesis, evaluation and deployment. You’ll work directly with clients, product owners and domain experts, and set technical direction for the engineers around you.

Relocation: This role requires relocating to the Cayman Islands within 3 months of joining.

What You’ll Do

- Build full stack, end to end: backend APIs, data persistence, and web interfaces such as chat and call dashboards, transcripts and agent configuration screens
- Lead voice agent delivery: real-time audio streaming, STT, LLM-driven dialogue, TTS, and telephony or WebRTC integration
- Engineer natural conversation: low-latency streaming, voice activity detection, barge-in handling, turn-taking, endpointing, and graceful escalation to a human
- Build agentic workflows with tool calling, multi-step planning, function orchestration and RAG
- Own evaluation: offline and online harnesses measuring latency, WER, task completion, groundedness and cost, with golden datasets and regression gates in CI




- Take systems to production: latency and cost optimisation, observability, call tracing, rate limiting, and failover across speech and model providers
- Handle sensitive data correctly: PII redaction across text, audio and transcripts, call-recording consent and retention, secrets management and RBAC
- Mentor engineers, review designs and code, set standards for prompting, evals and LLMOps, and contribute to client architecture reviews and estimates

Tech You’ll Work With

- LLMs & Orchestration: Azure OpenAI Service, OpenAI Realtime API, Semantic Kernel, LangChain/LangGraph, LlamaIndex, Autogen, Pipecat, MCP
- Voice & Speech: Azure AI Speech, Azure AI Foundry voice models, Azure Communication Services, Deepgram, ElevenLabs, Twilio, LiveKit, Vapi, Retell
- Telephony & Contact Centre: SIP, WebRTC, WebSockets, Genesys, Five9, Amazon Connect, Twilio Flex
- Retrieval & Data: Azure AI Search, pgvector, Pinecone, Qdrant, Azure Document Intelligence, relational and document stores
- Backend: Python (FastAPI), .NET / C# (ASP.NET Core), async programming
- Frontend: React, Angular or Vue with TypeScript
- Cloud & DevOps: Azure, Docker, AKS, Azure Container Apps, Azure DevOps, GitHub Actions, Bicep, Terraform
- Security: OAuth 2.0, RBAC, PII redaction, secrets management

What We’re Looking For

- 8+ years of software engineering, including 3+ years of hands-on GenAI/LLM work on systems you designed, shipped and supported in production (not prototypes or internal demos)
- Strong backend skills in both Python and .NET (C#)




- Full-stack capability: a modern TypeScript front end, REST APIs, background jobs, auth, data stores and cloud deployment
- Experience with at least one orchestration framework (Semantic Kernel, LangGraph, LlamaIndex, Autogen or Pipecat)
- Hands-on RAG experience: embeddings, vector databases, hybrid retrieval, and citations and grounding
- Experience building voice agents (STT, TTS, real-time streaming, LLM orchestration), ideally with a shipped phone or web voice product
- Real-time communication experience with WebSockets and/or WebRTC
- A track record of measuring and improving LLM and voice systems quantitatively
- Strong fundamentals in API design, data modelling, testing, CI/CD and infrastructure as code
- Experience with sensitive data and security controls
- Transparent communication of trade-offs to non-technical stakeholders

Bonus Points

- Azure certifications (AI-102, AZ-305)
- Speech-to-speech and realtime multimodal models, or custom vocabulary and custom neural voice
- Multilingual, accent-robust voice solutions
- Contact centre, IVR or CCaaS integrations, including latency, jitter and codec handling
- Call analytics: summarisation, sentiment analysis, QA scoring
- Voice or LLM solutions built for regulated industries (healthcare, financial services, insurance)
- Multimodal or document-heavy RAG
- Client-facing delivery experience

Why You’ll Love It

- Your GenAI, voice and full-stack work ships to real end users
- You’ll be part of a small, senior team with direct access to leadership and clients
- You’ll use a modern cloud and AI stack, with budget to experiment with new language and voice models
- You’ll build on The Protean, our in-house AI platform with 100+ agents
- You’ll be based in the Cayman Islands

Winter Inc. is an equal opportunity employer. We welcome applicants of all backgrounds.

📌 Senior AI Engineer (India)
🏢 Winter
📍 India