AI Data Engineer (Speech & Language)
Blue Machines (company site, opens in a new tab)1–2 yrsBengaluruOn-siteDirect career pageEst. 6–12 LPA
Posted 3d ago
Reach people at Blue Machines
People at this company you can reach
Log in to see who works here
Create a free account or log in to see the people at Blue Machines you could write to:
- Recruiters & HR
- People in this function
- Work addresses
Your match
Against your primary resume
Log in to see your match score
Create a free account or log in to unlock your CV match score across:
- Role fit
- Experience
- Industry relevance
Protocol Intelligence
How competitive this role is
Competition
Background fit
Strictness
Log in to see why each signal reads the way it does.
Job description
Structured overview of role & requirements
About This Role
- Own data preparation across pretraining, supervised fine-tuning and reinforcement learning for speech and language models serving Indian languages and code-mixed users.
- Build end-to-end data pipelines: source, clean, deduplicate, redact and track licensing; generate synthetic text/audio; package versioned, reproducible training datasets.
- Create language- and domain-specific evaluation sets and metrics, then use production voice-agent failures to improve training and test data.
Minimum Requirements
- Work Experience Required: 1–2 years building data pipelines, ML pipelines or backend systems; strong internships or open-source work count.
- Strong Python; Linux, Bash and Git; hands-on use of Hugging Face Datasets, pandas or Polars, plus audio tools such as librosa, torchaudio, sox or ffmpeg.
- Working knowledge of training STT, TTS, LLM and classifier models; experience with parallel or distributed jobs and cloud object storage.
- Familiarity with Indian languages and scripts, and ability to read or speak at least one Indian language besides English.
Ideal Candidate Profile
- Has built practical data or ML pipelines and can take datasets from collection through quality checks, model-ready packaging and reproducible runs.
- Can work across speech and language model workflows, including evaluation and production-feedback loops, rather than focusing on a single model stage.
- Has strong Indic-language fit, especially for script, Unicode, transliteration and code-mixed data quality.