MLOps Engineer

Job Description

About the Role:

We are seeking an MLOps Engineer who can deploy, optimize, and maintain AI models particularly LLMs and Voice AI in real-world environments.
The main focus of this role is on deploying language models (such as DeepSeek, OpenAI API), working with local GPUs or dedicated servers, and managing real-time communication with language and voice-to-text services.

Responsibilities:

Work with APIs of large language models (OpenAI, DeepSeek, and similar), including managing API keys, rate limits, and stable connections.

Install, configure, and deploy LLMs on GPUs (e.g., DeepSeek, Mistral, Llama, etc.).

Implement and integrate Voice-to-Text solutions (such as Whisper or Google Speech API).

Create and maintain streaming connections to LLMs for live and real-time responses.

Monitor GPU usage, RAM consumption, and task loads, and optimize system performance.

Write scripts for simple automation of deployment or monitoring (using Python or Bash).

Collaborate closely with the backend and model teams to ensure smooth and stable system performance.

Required Skills and Experience:

Proficiency in Python and basic ML libraries (PyTorch or TensorFlow, at least for execution and configuration).

Hands-on experience deploying LLMs on GPUs.

Familiarity with APIs such as OpenAI, DeepSeek, and similar services.

Good understanding of GPU operations (nvidia-smi, memory usage, batching, etc.).

Experience with lightweight monitoring tools such as Prometheus or Grafana (basic level).

Ability to work in Linux environments and familiarity with Docker for simple deployments.

Knowledge of Voice-to-Text frameworks such as Whisper, Vosk, or SpeechRecognition.

Nice to Haves:

Experience with streaming or WebSocket connections to models.

Familiarity with LLM quantization or optimization techniques.

Interest in inference optimization and latency reduction.