We are building a strong DevOps, MLOps, and DataOps team to test, validate, and deploy production-grade Computer Vision and AI pipelines. The team's mission is to handle pipelines running on a wide range of hardware, from milliwatt edge devices like Jetson Orin and Hailo-8/15 to multi-GPU data-center nodes. As a member of this team, you will need to master DevOps, MLOps, and DataOps while maintaining hardware awareness. This is a senior-level position, and you will report to the CTO.

Skills

• DevOps
• MLOps
• Performance Testing
• Docker/Kubernates
• GIT
• Willingness to Learn

Requirements

• Job Role MLOps Engineer
• Job Type Full Time
• Workplace Type Onsite
• Industry Information Technology & Services

Secondary locations

About

Not provided

Responsibilities

As a junior member of the team, you will support the process of moving AI/ML artifacts from R&D to production.  Your key contributions will involve:
* Assisting with functional testing, including unit, integration, and regression tests.
* Helping to validate models against key metrics like latency, throughput, and accuracy. Supporting the path to production through tasks related to containerization, CI/CD, and monitoring.

Role Focus and Time Allocation

Focus: Your work will be centered on edge testing, edge deployment, automated regression, and release operations.

Time Allocation:

•Testing: 60%

•Deployment & Ops: 40%

*Responsibilities*

In this entry-level role, you will learn and apply the following, with guidance from senior team members:

*Build & Packaging:*

Edge: Assist with tools like TensorRT, ONNX, and Hailo NN-Converter.

Server / Data-Center: Work with Yocto / Ubuntu Core images.

Cross-Cutting: Use Docker/OCI containers, help package testing environments, and work with Git-versioned artifacts.

Functional & Model Validation:

•Edge: Support model quantization and calibration tasks.

•Server / Data-Center: Help run fairness tests.

•Cross-Cutting: Work with regression datasets in DVC.

Performance & Power Testing:

•Edge: Use tools for Jetson power-profiling and assist in thermal testing.

•Server / Data-Center: Support bandwidth and saturation tests.

•Cross-Cutting: Help measure and report latency and throughput.

Other Requirements

We are looking for candidates with a foundational skill set who are eager to learn. Basic familiarity with performance testing and Docker/Kubernetes deployments is required.

Modern C++ (17/20/23):

•Basic CMake skills and familiarity with gdb.

AI Runtimes – C++ APIs:

•ONNX Runtime (C++): Ability to run sample applications.

•TensorRT (C++): Experience with trtexec.

•NVIDIA Triton: Capable of deploying a prebuilt model.

•DeepStream (C/C++): Ability to run the reference application.

Edge & Embedded Tool-chain:

•Jetson (L4T): Proficient with ssh and tegrastats.

•Hailo-8/15: Able to run samples.

•Intel ARC / OpenVINO: Can run basic hello_query_device.

Build & Packaging (C++):

•CMake: Can build from a given README file.

•Conan / vcpkg: Able to install dependencies.

•Cross-compilation: Can use prebuilt images.

Performance & Profiling:

•CPU: Familiar with htop.

•GPU: Familiar with nvidia-smi.

Testing C++ Pipelines:

•Unit / Integration: Can write simple assert statements for testing.

•Regression & KPI: Able to run scripts, collate data and KPIs, and report them to the R&D team.

DataOps & Telemetry:

•gRPC / ZeroMQ C++: Can run example code.

•Logging & Tracing: Can use printf for logging.

Experience

• 3 -5 years of experience.

• Solid C++ coursework or open-source contributions (please provide proof via GitHub links).

Good to have

Not provided

About the Company

Trois Infotech is a cutting edge technology company having products and solutions for Responsible surveillance, Cleantech, Telecommunication and Utility. Our state-of-the-art AI and computer vision research team is committed to transform the camera-based surveillance. Our research includes edge and server-based computer vision products for security surveillance, traffic enforcement, defense, law and order, crime analysis, perimeter security with violations and incident detection, unconstrained ANPR, Pose invariant Face Recognition, Gait Recognition etc. In the Cleantech domain, we have an enterprise waste management product which can track and monitor the entire waste management process with role-based web and mobile applications and insightful reports and dashboards. We have large scale wide implementation of computer vision products and platform and mobile applications for waste management monitoring. We have ongoing projects in India and the Middle East. The current total team size is more than 100.