Software Engineer II – Machine Learning

Position Overview

We are seeking a Software Engineer II (ML Engineer) to own, maintain, and scale production deep-learning inference services and evaluation pipelines for Perceptual Audio Evaluation. In this role, you will manage always-on ML inference capacity, integrate models into internal toolsets and lightweight web UIs, execute model evaluations, and communicate directly with audio engineers, research scientists, and cross-functional teams.

Key Responsibilities & Deliverables

• ML Model Ownership & Operations: Own a family of deep-learning models end-to-end (architecture, checkpoints, evaluation pipelines, serving infrastructure, and failure modes).

• Inference Capacity & Monitoring: Operate always-on model inference capacity—monitoring traffic, resolving throttling, tuning auto-scaling rules, requesting capacity, and redeploying endpoints.

• Tool & API Integration: Integrate ML models into internal and cross-functional workflows via REST/GraphQL endpoints and lightweight web UIs.

• Evaluations & Minor Fixes: Run model evaluations on request, apply preprocessing updates, fix minor bugs, and manage version bumps/checkpoint swaps.

• On-Call & User Support: Serve as on-call support for covered services, addressing ticket queues, running runbooks, and providing technical support to audio engineers, SDEs, research scientists, and TPMs.

Required Qualifications & Skills

• Education: Bachelor’s degree in Computer Science, Electrical Engineering, Audio Engineering, or a related technical field.

• Programming & ML Frameworks: Strong proficiency in Python and deep-learning frameworks such as PyTorch.

• ML Concepts: Solid understanding of Machine Learning concepts, inference serving, and ML engineering practices.

• Audio Fundamentals: Foundational understanding of audio and signal processing concepts (waveforms, sample rate, spectrograms) to evaluate and sanity-check model outputs.

Preferred Qualifications

• Master's or PhD in Electrical/Audio Engineering, Speech/Signal Processing, Acoustics, or Computer Science.

• 2+ years of hands-on experience deploying, serving, and maintaining production ML models (including on-call, runbooks, and incident response).

• Experience with audio/speech/perceptual quality models (e.g., MOS prediction).

• Familiarity with Meta’s internal ML platform tools (Bento, internal model serving infrastructure).

• Experience building lightweight web UIs or front-end onboarding flows.