Leads the team this role joins
Heads the group this role sits in
In-house, owns the opening
From your own Gmail and LinkedIn.
About the role
This is a product ML engineering role at Adalat AI for engineers in India with 2-6 years of production ML experience, and it can be done remotely from India. The job is about turning research-grade models into systems that hold up in courtrooms, covering search, retrieval, legal reasoning, speech, and translation for judges, lawyers, and stenographers. The engineer also builds evaluation infrastructure and feedback loops that carry real-world failures back to the research team.
What you’ll do
- Build retrieval and search agents across multilingual, multi-document legal collections
- Develop legal reasoning pipelines that help users work through case history, orders, and filings
- Build production speech systems for transcription, speaker diarization, and voice-to-action workflows
- Own translation and summarization pipelines for legal proceedings in Indian languages
- Design evaluation harnesses that measure what matters in a courtroom setting
- Build feedback pipelines that surface real-world model failures to researchers
- Turn courtroom audio, transcripts, and documents into training signal, and design annotation pipelines with the Data Operations team
- Take research prototypes to production and flag distribution shifts and edge cases before they become incidents
What they’re looking for
- 2-6 years of building ML systems in production
- Strong Python and familiarity with the modern ML stack, including PyTorch and Hugging Face
- Hands-on experience with at least one of: LLM fine-tuning and deployment, ASR or speech pipelines, RAG and retrieval systems, or NLP information extraction
- Experience designing and running model evaluations, deciding what to measure rather than only running benchmarks
- Experience shipping and maintaining production code, not only prototypes
Nice to have
- Deep specialization in speech (ASR, diarization, multilingual audio) or text (LLMs, retrieval, legal NLP)
- Experience with Indic languages
- Prior work in legal, healthcare, government, or civic tech
- Publications at ML venues
Skills
Questions about this role
Is this Adalat AI ML Engineer role remote?
The listed location is remote within India.
How many years of experience are needed?
2-6 years of building ML systems in production.
What is the hiring process?
Five rounds: an intro call, an ML team deep-dive, a take-home assignment review, a culture fit chat with a co-founder, and then an offer if it is a match.
Do I need a speech or text specialization?
You can come in with a specialization in either text or speech, or as a generalist who works across both.