DevOps Engineer

DevOps Engineer, AI Voice Infrastructure
Mumbai · Full-time · In-office · Andheri West
Scale AI that takes real restaurant orders
About Bundl AI
Bundl AI builds AI-native operating systems for physical businesses.
We’re building Bundl OS, software that brings business operations, workflows and AI employees together. Our AI employees don’t just answer questions. They interact with customers, work with business systems and take actions in the real world.
One of our key products is an AI restaurant order taker that answers incoming phone calls, understands what customers want, handles menu and availability questions, captures orders and connects the conversation to restaurant operations.
We’re building infrastructure to take this from individual deployments to a platform capable of handling thousands of restaurant calls reliably.
We’re an early-stage, high-growth company building from Mumbai for global markets.
The role
We’re looking for a DevOps Engineer focused on scaling real-time AI voice systems in production.
Your primary responsibility will be to make our AI restaurant caller reliable, responsive and cost-efficient as call volume grows.
You’ll work across cloud infrastructure, telephony, real-time audio, AI services, networking and deployment automation. You’ll identify bottlenecks, design scaling strategies, improve monitoring and ensure the system continues working under heavy concurrent traffic.
This is a hands-on engineering role. We need someone who can investigate a production issue, understand the entire call path, find the bottleneck and build a durable solution.
What you’ll own
Scaling and concurrency
* Design infrastructure to handle increasing numbers of simultaneous restaurant calls.
* Build and tune autoscaling for real-time voice workloads.
* Load-test the complete call flow and identify concurrency limits.
* Prevent overload through backpressure, queuing, rate limits and graceful degradation.
* Improve resource utilisation across compute, audio processing and AI services.
* Plan capacity and reliability as the number of restaurants and call volumes grows.
Real-time voice infrastructure
* Operate and optimise the infrastructure connecting telephony, audio streams, speech-to-text, LLMs and text-to-speech.
* Improve end-to-end latency so conversations feel natural and responsive.
* Debug dropped calls, audio interruptions, connection failures and degraded call quality.
* Work with SIP, telephony providers, Asterisk, WebRTC and LiveKit where applicable.
* Build resilience around external AI and telephony provider failures.
* Ensure concurrent calls remain isolated and one customer’s traffic does not disrupt another’s.
Cloud and deployment infrastructure
* Build and maintain AWS infrastructure for production voice and backend services.
* Improve Docker-based deployments and Kubernetes-based orchestration where appropriate.
* Build CI/CD pipelines for safe, repeatable releases.
* Automate provisioning and deployment of services across environments.
* Manage networking, load balancing, secrets, access controls and backups.
* Establish reliable staging environments for testing new infrastructure and voice-system changes.
Observability and reliability
* Implement metrics, logs, traces, dashboards and actionable alerts.
* Monitor concurrent calls, call failures, connection times, response latency and infrastructure utilisation.
* Build tools to investigate individual calls and diagnose failures across the entire pipeline.
* Define and track uptime, latency and call-completion targets.
* Develop incident response, recovery and rollback procedures.
* Identify failure patterns before they affect restaurant customers.
Cost and performance
* Track infrastructure and AI-provider costs per call and per minute.
* Optimise compute, networking, audio processing and service utilisation.
* Identify opportunities to reduce idle capacity without compromising call quality.
* Benchmark infrastructure choices against throughput, latency, reliability and cost.
* Help establish the economics required to scale the product across many restaurants.
What we’re looking for
* Experience operating and scaling production services.
* Strong Linux, networking, DNS, HTTP and troubleshooting fundamentals.
* Hands-on AWS or equivalent cloud infrastructure experience.
* Experience with Docker, deployment automation and CI/CD.
* Understanding of concurrency, resource management and distributed systems.
* Experience with monitoring, logging and production incident resolution.
* Scripting skills in Bash, Python or similar languages.
* Ability to diagnose complex problems across multiple connected services.
* A strong ownership mindset and willingness to work hands-on with production systems.
Particularly valuable experience
* Kubernetes, autoscaling and container orchestration.
* Real-time communication infrastructure such as WebRTC or LiveKit.
* SIP, Asterisk, telephony gateways or programmable voice APIs.
* Streaming audio pipelines and speech-to-text/text-to-speech services.
* High-concurrency systems, event-driven architectures and asynchronous processing.
* Load testing, capacity planning and performance engineering.
* Production AI agents, LLM APIs or latency-sensitive inference workloads.
* Experience scaling voice bots, contact-centre platforms or conversational AI products.
You don’t need experience with every tool listed. Strong infrastructure fundamentals and a track record of solving production scaling problems matter most.
What success looks like
* More simultaneous calls without a proportional increase in operational complexity.
* Low and predictable conversational latency under load.
* Fewer dropped calls and production failures.
* Reliable deployments with minimal disruption to active calls.
* Clear visibility into bottlenecks, incidents and capacity.
* Lower infrastructure cost per completed call.
* A platform that can onboard more restaurants without repeatedly rebuilding the infrastructure.
How we work
You’ll work directly with the founders and engineering team on a live product used by real businesses.
You’ll have meaningful ownership of architecture, infrastructure and production reliability. You’ll also work closely with application and AI engineers to improve the system as a whole.
We move quickly and expect engineers to investigate problems, test hypotheses and ship practical solutions.
Why Bundl?
* Work on a real-time AI system with tangible commercial use.
* Solve difficult problems at the intersection of telephony, AI and distributed systems.
* Own infrastructure that directly affects customer experience and unit economics.
* Help scale a product from individual deployments to a multi-customer platform.
* Work directly with the founders and build from Mumbai for global markets.
* Participate in ESOPs as an early member of the team.
Compensation
Competitive salary based on experience, with ESOP participation.
Location
Andheri West, Mumbai · Full-time · Fully in-office
The challenge
How do we make an AI restaurant order taker handle more calls, respond faster and remain reliable as demand scales?
If you enjoy solving real production engineering problems, we’d like to meet you.

Work arrangement
No