Machine Learning Engineer

About Good Tape

Good Tape is building tools that make working with voice and audio effortless. We help journalists, researchers, and organizations capture, transcribe, and analyze conversations at scale, supporting 100+ languages with AI-powered features like semantic search, summarization, and speaker diarization. Our team is small, ambitious, and deeply technical.

We care about reliability, craft, and shipping things that solve real problems for our users.

The role:

We are looking for a Machine Learning Engineer to own our ML stack, from fine-tuning ASR models to designing multi-model inference architectures and deploying them in production.

You will be working on the core of what makes Good Tape good: the accuracy, speed, and reliability of our transcription. This is a hands-on role with real ownership. The work you ship is immediately felt by tens of thousands of users.

What you'll work on:

• Our ASR pipelines — fine-tuning models, designing multi-model inference architectures, and improving speed, accuracy, and reliability across 100+ languages

• Diarization and speaker attribution

• Translation, NER, and reranking pipelines

• Real-time transcription — we have a PoC in progress and need someone to help take it to production

• Model evaluation, benchmarking, and quality tracking across languages

What We’re Looking For:

• Strong Python — you will be writing production FastAPI services, not notebooks

• Deep experience in the ASR domain — you know the landscape (Whisper, wav2vec, Conformer-based models) and have fine-tuned models on custom data

• Comfortable with Docker and Kubernetes — you deploy your own work

• You have shipped models to production, not just trained them

Nice to have: general NLP understanding (classification, NER, summarization), experience with CTranslate2 or vLLM, speech-specific models (pyannote, NeMo), distributed inference with Redis.

Why join Good Tape?

• Work on a product that makes real impact for journalists, researchers, and teams worldwide

• Small, focused team — your work has direct and visible impact

• Modern AI infrastructure: multi-model GPU pipelines, multi-cluster Kubernetes across GCP and Scaleway

• Real ownership and agency — we care about good engineering, not process for the sake of it.

• Small team where your work is visible and your input shapes the product