This is a pure Individual Contributor role within our R& D function, working across STT, LLM, and TTS systems. The mandate is the same one every research role here carries: close the distance between research and production. You'll work directly with backend engineers to take your work from a working prototype to something running reliably in front of real customers, with the observability it needs to be trusted at enterprise scale.
Responsibilities
- Improve STT/ASR transcription accuracy and reduce latency on specific domain and language targets, within multilingual, financial-services voice data.
- Fine-tune, evaluate, and deploy LLMs for defined BFSI tasks: information extraction, classification, summarisation, and compliance signal detection.
- Build and benchmark TTS improvements against product requirements - quality, naturalness, latency, integration fit.
- Extend and maintain our prompt engineering and RAG infrastructure for production LLM features.
- Work directly with backend engineers to take your research output from prototype to deployed, observable production feature.
- Run experiments and contribute improvements to our evaluation frameworks, so results are reproducible and tied to real product outcomes.
- Track developments in open-source LLM and ASR frameworks and bring evidence-backed recommendations to the team.
Requirements
- 5-8 years in software engineering, with meaningful depth in ML/NLP systems.
- Hands-on experience with LLMs - from prompt design through fine-tuning, evaluation, and deployment.
- Exposure to ASR/STT technologies: Whisper, Kaldi, DeepSpeech, or commercial equivalents.
- Proficiency with ML tooling: Hugging Face, LangChain, or equivalent frameworks.
- Cloud experience (AWS or GCP) for model training, deployment, and monitoring.
- Comfortable reasoning through modelling and architecture trade-offs with incomplete information, and can explain that reasoning clearly to the wider engineering team.
- Writes clean, production-ready Python that backend engineers can integrate and maintain.
- Understands how AI/ML components fit into larger backend architectures.
Strong Signals
- Has closed the gap between this works in a notebook and this is running reliably in production.
- Has worked directly with backend engineers to ship an AI-powered feature.
- Holds a high bar on evaluation.
- Can take a scoped but underspecified problem and turn it into a working plan.
- Can explain a technical trade-off or limitation to a product or business stakeholder without losing precision.
This job was posted by Aishwarya Dsouza from GreyLabs AI.
