Research Engineer

Revoize Inc. is a deep-tech company building real-time, on-device speech-to-speech translation (S2ST) that preserves the speaker's voice and emotion. The company is hiring proven researchers to build a research team and reach on-device S2ST models within the next 12 months.

The Research Engineer will work on the core speech translation models (training, evaluation and robustness) as part of the research team. Reports to the Research Manager once hired; until then to the CEO.

Key information:

Salary:

• 6 000 - 9 500 USD net/month - pure B2B contract

Wymagania

It is vital that you have:

• PhD in ML / speech / NLP (fresh graduates with strong publications welcome), or MSc with a strong research record.

• 3+ years of hands-on ML research in speech, audio or NLP; PhD years count.

• Hands-on training of ASR, speech-to-text, machine translation or speech generation models at scale, from data to evaluation.

• Solid PyTorch; runs own experiments end to end, including ablations and error analysis.

• Good understanding of transformers / sequence-to-sequence and language model training.

• Publications or other verifiable track record of original work.

• Independent, startup mindset;

• fluent English (min. C1).

We will be more than happy if you have:

• Neural audio codecs;

• speech language models;

• streaming / simultaneous translation;

• quantization and distillation;

• on-device inference;

• reinforcement learning;

• multilingual work (Polish a plus).

Technology stack:

• Python, PyTorch;

• transformers / sequence-to-sequence models;

• ASR and machine translation;

• speech and audio processing;

• multi-GPU training;

• TensorBoard;

• Linux;

• Git;

• Docker.

Obowiązki

You will be responsible for supporting our team in:

• develop, train and evaluate speech translation models (speech-to-text, text-to-text, speech-to-speech), including streaming operation.

• run experiments and ablations; analyse results and report them to the Research Manager.

• improve robustness to real-world conditions (noise, accents, conversational speech).

• work with the data team on training data quality and coverage.

• work with the Research Manager and MLOps on latency and model size for on-device deployment.

• keep experiments documented and reproducible.

Oferujemy

What do we offer:

• work 100% remote;

• b2b contract;

• in addition to the cash salary, Revoize offers an ESOP (employee stock option plan);

• Claude Max – Extended Use (Max 5x) or Claude Max – Max Flexibility (Max 20x).