Build voice AI for India
We’re hiring engineers and designers at Nupaka Industries to ship real phone agents — ML, Python, OOP, and calm product UX for enterprise customers.
ML Engineer — Speech & NLP
The role
Own STT/TTS quality for Indian languages — latency, code-switching, and telephony-grade audio pipelines on our voice platform.
Who should apply
B.E./B.Tech or M.Tech in CS/EE with 2+ years in speech or NLP. Comfortable with Python, PyTorch or similar, and shipping models to production.
Expectations
- Fine-tune and evaluate ASR/TTS for Hindi, Telugu, Tamil, and English code-mix
- Build offline eval sets and regression tests for model releases
- Collaborate with backend on streaming inference at 8–16 kHz
- Document model cards, failure modes, and language-specific tuning
Senior ML Engineer — LLM & Voice Agents
The role
Design agent behaviour — prompts, tool use, turn-taking, and safety — for outbound sales and support calls in Indian languages.
Who should apply
5+ years in ML/NLP with production LLM experience. You’ve shipped conversational systems, not just notebooks.
Expectations
- Own industry-specific agent scripts (real estate, NBFC, edtech)
- Reduce hallucinations and enforce business rules in voice replies
- Measure quality: WER, task completion, and human review loops
- Mentor junior ML engineers on the team
MLOps Engineer
The role
Run model serving, observability, and CI for STT/LLM/TTS — from dev laptops to EC2 and GPU instances.
Who should apply
3+ years in DevOps/MLOps with strong Python and Linux. Experience with Docker, metrics, and incident response.
Expectations
- Containerise inference services and manage rollouts with zero-downtime
- Track latency, GPU utilisation, and cost per call minute
- Automate dataset versioning and model promotion gates
- On-call rotation for production voice traffic
UI/UX Product Designer
The role
Shape the CRM dashboard, call review flows, and marketing site — calm enterprise UX, not generic SaaS templates.
Who should apply
3+ years product design with a strong portfolio in B2B or fintech. Figma expert; HTML/CSS awareness helpful.
Expectations
- End-to-end flows: wireframes → high-fidelity → dev handoff
- Design system tokens, typography, and accessible components
- Partner with engineering on live dashboard and call-detail UX
- Present work to founders and enterprise customers
Senior UX Designer — Voice & Conversation
The role
Design how agents speak, hand off, and recover errors — conversation maps, tone guidelines, and multilingual UX patterns.
Who should apply
5+ years in UX with voice, chatbot, or IVR experience. You understand Indian language switching and enterprise buyers.
Expectations
- Write conversation scripts and review live call transcripts
- Define UX metrics: completion rate, escalation, customer satisfaction
- Run usability tests on demo calls with real users
- Align visual brand with voice persona across industries
UX Researcher
The role
Validate product decisions with sales teams, call-centre leads, and end customers across India’s language markets.
Who should apply
2+ years in UX research or market research. Fluent in English plus at least one Indian language. Strong interview skills.
Expectations
- Plan and run interviews, diary studies, and call listening sessions
- Synthesise insights into actionable product briefs
- Track competitive landscape in voice AI and CRM tools
- Support sales with customer quotes and case-study material
Python Backend Engineer
The role
Build FastAPI services — WebSockets, async Postgres, contact APIs, and integrations for our voice orchestrator.
Who should apply
2+ years Python backend. Solid async/await, REST, and SQL. B.E./B.Tech or equivalent practical experience.
Expectations
- Implement and test API endpoints with asyncpg and pydantic
- Handle WebSocket sessions at scale with clean session isolation
- Write readable code, unit tests, and API documentation
- Deploy to AWS EC2 with nginx and systemd
Senior Python Engineer — Voice Platform
The role
Architect the real-time pipeline: STT → LLM → TTS with backpressure, barge-in, and per-call state management.
Who should apply
5+ years Python with production async systems. You’ve built low-latency services and care about clean abstractions.
Expectations
- Own core modules in main orchestrator and refactor for testability
- Profile and optimise end-to-end call latency
- Design OOP boundaries between media, AI, and persistence layers
- Review code and set engineering standards for the team
Software Engineer — OOP & Systems
The role
Build reliable subsystems — session managers, audio codecs, dialplan adapters — with clear class design and interfaces.
Who should apply
3+ years in C++, Python, or Java with strong OOP fundamentals. CS degree or proven systems background.
Expectations
- Apply SOLID principles to voice-session and media-stream code
- Integrate FreeSWITCH/SIP components when we go telephony-live
- Write integration tests for call lifecycle edge cases
- Document architecture decisions for the team
Backend Engineer — Object-Oriented Architecture
The role
Model domain entities — calls, leads, agents, industries — and keep business logic out of HTTP handlers.
Who should apply
3+ years backend development with explicit domain modelling experience. Patterns: repository, factory, strategy.
Expectations
- Refactor monolith modules into testable domain services
- Define interfaces for CRM, billing, and webhook integrations
- Enforce typing and linting in the Python codebase
- Pair with ML team on clean API contracts
Full Stack Engineer — Python & Frontend
The role
Ship customer-facing dashboards and demo experiences — Python APIs plus polished HTML/JS interfaces.
Who should apply
3+ years full stack. Strong Python plus modern CSS/JS. Eye for detail matching our calm enterprise aesthetic.
Expectations
- Build dashboard, careers, and marketing pages with design fidelity
- Wire WebSocket demos to backend with minimal latency
- Implement responsive layouts and accessible interactions
- Collaborate with UX on prototypes and user testing
Data Scientist — Call Analytics
The role
Turn call transcripts into outcomes — site visits booked, hot leads, PTP — and dashboards customers trust.
Who should apply
2+ years in data science or applied ML. Python, pandas, SQL, and classification/NLP basics. Stats foundation required.
Expectations
- Label and model outcome detection from multilingual transcripts
- Build reporting for ops teams: volume, conversion, language mix
- Work with product on CRM auto-fill and lead scoring
- Ensure privacy and consent in data handling