Experience: 3–7 Years
Domain: AI/ML, Edge Computing, Video Surveillance & Defence Technology
We are looking for an experienced DevOps Engineer to manage the packaging, deployment, monitoring, and maintenance of real-time EO/IR (Electro-Optical/Infrared) vision systems across edge GPU nodes and central servers.
The role involves building reliable and repeatable deployment pipelines for containerized applications, AI inference runtimes, GPU environments, and supporting backend services within an integrated surveillance and command system.
The ideal candidate will have hands-on experience with Docker, CI/CD, Linux systems, NVIDIA GPU runtimes, CUDA/TensorRT compatibility, and production release management. Experience deploying AI applications on edge devices or working in restricted-network environments will be a strong advantage.
Key Responsibilities
- Build, maintain, and optimize CI/CD pipelines for frontend applications, APIs, Python AI runtimes, and AI model/engine artifacts.
- Develop and manage Docker images for x86_64 GPU servers and ARM64/aarch64 edge devices, including NVIDIA Jetson platforms.
- Manage Docker Compose-based deployments for services such as APIs, LiveKit/SFU, PostgreSQL/PostGIS, MinIO, and map tile services.
- Maintain compatible NVIDIA CUDA, TensorRT, GPU driver, NVIDIA Container Toolkit, and JetPack runtime configurations.
- Manage staged releases across development labs, staging environments, and customer sites.
- Validate deployment health, service availability, GPU utilization, application logs, and system performance after updates.
- Implement release versioning, configuration management, model-pack promotion, and rollback procedures.
- Troubleshoot production issues involving video streams, GPU drivers, AI engine loading, disk capacity, containers, and service failures.
- Manage secrets, environment variables, credentials, ports, and site-specific configuration files securely.
- Maintain backup and recovery procedures for PostgreSQL/PostGIS databases and S3-compatible storage such as MinIO.
- Develop deployment runbooks, incident response procedures, release checklists, and technical documentation.
- Coordinate deployment schedules and maintenance windows with customer IT teams while adhering to network restrictions and access-control policies.
- Maintain reproducible deployment baselines using container image digests, model/engine checksums, and configuration manifests.
- Support resource monitoring and optimization for concurrent video streams, AI inference, tracking, mapping, and API services.
Required Skills & Technical Experience
- Experience: 3–7 years in DevOps, Platform Engineering, SRE, or production deployment of containerized applications.
- Docker & Containers: Strong experience with Docker, multi-stage builds, Docker Compose, container registries, and image versioning.
- CI/CD: Hands-on experience with GitHub Actions, GitLab CI/CD, Azure DevOps, or equivalent pipeline tools.
- Linux Administration: Experience with Linux services, systemd, shell scripting, logs, disk management, GPU monitoring, and troubleshooting.
- NVIDIA GPU Stack: Practical experience managing CUDA, TensorRT, NVIDIA drivers, NVIDIA Container Toolkit, and runtime compatibility.
- Edge Deployment: Experience building and deploying applications for x86_64 GPU servers and ARM64/aarch64 devices. Familiarity with NVIDIA Jetson and JetPack is highly desirable.
- Release Engineering: Experience managing versioned application releases, AI model artifacts, engine binaries, configuration files, and rollback workflows.
- Configuration & Security: Knowledge of environment variables, secrets management, secure credential handling, and controlled configuration changes.
- Storage & Databases: Familiarity with PostgreSQL, PostGIS, S3-compatible storage, MinIO, and backup/restore procedures.
- Monitoring & Recovery: Ability to troubleshoot production incidents and implement reliable health checks, monitoring, and recovery processes.
Preferred Qualifications
- B.E./B.Tech in Computer Science, Information Technology, Electronics & Communication, or a related discipline; equivalent practical experience may be considered.
- Experience deploying computer vision, AI inference, or video analytics applications on edge GPU hardware.
- Familiarity with NVIDIA Jetson flashing, image creation, and device provisioning.
- Experience with Ansible, Terraform, Prometheus, Grafana, and infrastructure automation.
- Knowledge of air-gapped deployments, offline package repositories, private registries, and low-bandwidth update strategies.
- Familiarity with LiveKit, SFU architecture, TURN networking, and real-time video streaming infrastructure.
- Experience working with defence, surveillance, industrial, or other security-sensitive systems.
Key Requirements for This Role
- Ability to reproduce production environments using versioned deployment manifests and artifact checksums.
- Strong understanding of GPU runtime compatibility and troubleshooting across development and production environments.
- Ability to execute controlled updates and recover failed deployments with minimal downtime.
- Comfort working on-site with restricted networks, customer IT teams, and scheduled maintenance windows.
- Strong ownership of production reliability, deployment quality, documentation, and incident resolution.
Interview Assessment
Candidates should be prepared to discuss:
- An end-to-end CI/CD pipeline they have designed, implemented, and maintained.
- A production incident they resolved and the steps taken to restore service.
- A rollback of an incompatible GPU runtime, container image, or AI model/engine release.
- Their approach to deploying containerized AI applications in restricted or offline environments.
- Sample deployment runbooks, release checklists, or production troubleshooting workflows, where available and appropriate to share.
Pay: ₹1,000,000.00 - ₹1,200,000.00 per year
Work Location: In person
