Backend Engineering Intern, Voice AI Platform

Zudu AI; Chennai District; On-site; INR 400,000 - 700,000; Full time

Benefits offered by this job

Mentorship

Full-time path

Job summary

Zudu AI in Chennai is building a real-time Voice AI platform for enterprise customers. You will ship to production within your first month, own small services end-to-end, and debug live call flows with the core team.

You will work on backend services in Python/FastAPI and Java/Spring Boot, integrate LLMs and speech tech, and gain production experience under mentorship toward a full-time path. We value hands-on projects, daily use of AI coding tools, and willingness to learn across APIs, logs,

Qualifications

• Final-year CS/IT student or 2025/2026 graduate.
• Proven backend project deployed and used by others (internship, freelance, or personal).
• Experience with an LLM API, embeddings, or tool calling, and ability to explain failures.

Responsibilities

• Build and maintain backend services and APIs in Python/FastAPI (primary) and Java/Spring Boot.
• Support real-time voice agent workflows on LiveKit, including SIP/telephony integration and call control.
• Integrate LLMs, speech (STT/TTS) and RAG into production agents.
• Debug across APIs, WebSocket and SIP flows, and Kubernetes workloads on GCP using logs, metrics and traces.
• Ship tested, reviewed code through CI/CD.

Skills

Final-year student

Backend project

LLM API

AI coding tools

HTTP/REST

Git

Stack traces

Python/Java

Education

Final-year CS/IT degree

Tools

FastAPI

Spring Boot

LiveKit

Job description

Zudu AI is an enterprise Voice AI platform. Our real-time AI voice agents handle live customer calls in production for enterprise customers in India and overseas. The stack: Python (FastAPI) and Java (Spring Boot) services, LiveKit for real-time media, SIP telephony, STT/TTS, LLMs, PostgreSQL, Qdrant, and Kubernetes on GCP.

The role

This is not a training programme. You will ship to production within your first month, own small services end to end, and debug live call flows with the core team. We build with AI tooling (Claude Code, Cursor) every day and expect you to.

What you'll do

• Build and maintain backend services and APIs in Python/FastAPI (primary) and Java/Spring Boot
• Build and support real-time voice agent workflows on LiveKit, including SIP/telephony integration and call control
• Integrate LLMs, speech (STT/TTS) and RAG (PostgreSQL + Qdrant) into production agents
• Debug across APIs, WebSocket and SIP flows, and Kubernetes workloads on GCP using logs, metrics and traces
• Ship tested, reviewed code through CI/CD

The bar (read before applying)

• Final-year student or 2025/2026 graduate in CS, IT or a related field
• At least one real backend project outside coursework (internship, freelance or personal): deployed, used by someone other than you, code on GitHub
• You have built something with an LLM API (RAG, embeddings, agents or tool calling) and can explain where it failed
• You use AI coding tools daily and know where their output needs fixing
• Solid fundamentals: HTTP/REST, SQL, async or concurrency basics, Git
• You can read a stack trace and logs and form a hypothesis before asking for help
• Comfortable in Python or Java, willing to work in both

Nice to have

• SIP, VoIP, WebRTC or LiveKit
• Open source contributions

What you'll gain

• Production experience on a real-time Voice AI platform (telephony, media pipelines, LLM integration, cloud-native infra), direct mentorship from the core engineering team, and a full-time path for strong performers.