We are seeking a skilled ML/AI Engineer to join our team and deliver cutting-edge ML/AI models in the ASR domain and legal insights more broadly. The core of this role is building, deploying, and serving ML models in production.
A key part of this role is speech engineering — prior ASR/speech experience is not a must, but a willingness to learn and grow into this domain is essential. Experience with LLM-based solutions and agentic flows is an advantage, not a requirement.
What you’ll do:
Build, deploy, and serve ML models in production — the core of the role — optimizing inference performance, latency, and efficient GPU utilization.
Support and extend our speech solutions, including 3rd-party ASR integrations and self-hosted ML models.
Evaluate and optimize key ASR/ML metrics such as WER, latency, F1 score, EER, and DER.
Secondary: Integrate LLMs and agentic flows into production solutions where they add value, including fine-tuning existing models.
What you’ll bring:
Core (required):
Solid background in machine/deep learning, with 4+ years of experience.
Strong Python skills and experience working in Linux environments.
Familiarity with deep learning libraries (e.g., PyTorch, TensorFlow) and training workflows.
Experience with MLOps pipelines and serving/deploying models in production.
Experience optimizing models for inference, including GPU acceleration.
Experience with Docker, gRPC, and serverless / micro-services architectures on cloud infrastructure (AWS/GCP/Azure).
Experience with real-time / streaming systems and low-latency processing (e.g. streaming pipelines, live sessions).
Nice to have (secondary / optional):
Experience building and productizing agentic flows and AI-driven solutions (1+ year).
Familiarity with LLM / agent orchestration frameworks such as CrewAI, AWS Bedrock, Temporal, AgentCore, or LiteLLM.
Prior ASR / speech experience — not required; willingness to learn this domain is what matters most.
