Senior Data Scientist - ASR
Eka.CareBengaluru, IndiaOn-Site
Needs a free Superjobs account. You'll finish on Eka.Care's site.
Skill score
Summary
This person trains and ships Eka Care's medical ASR models, from multimodal speech-LLM and streaming CTC/transducer work through to GPU serving inside EkaScribe, and owns the data and evaluation pipelines that decide readiness. It suits a senior ASR engineer with 2-6 years of production speech experience who is strong in PyTorch, multi-GPU training and evaluation design.
- Function
- Data Science
- Estimated salary
- Est. ₹11–16 LPA
- Estimated for a senior role, ~3 yrs experience, India. Not published by the employer.
- Work mode and location
- On-Site, Bengaluru, India
- Experience
- 3–6 yrs
Job description
Overview
You will own the next generations of Parrotlet-A, Eka Care’s open-source medical ASR models, and ship them into EkaScribe, which turns doctor-patient conversations across India into structured clinical notes. Indian clinics speak Hinglish, Tanglish and a dozen other code-mixed registers over phone-grade audio, with drug names no general-purpose ASR has heard. Parrotlet-A 2 Pro already beats the best general ASR models on varied benchmarks.
Your job: widen that gap, close the remaining one on drug-entity accuracy, and lead the shift to multimodal LLM-based ASR (speech-LLMs that transcribe, normalise and understand a consultation in one pass), while extending to new languages, noise conditions and clinical settings.
Roles and responsibilities
- Train the next Parrotlet-A models end-to-end. Multimodal LLM-based ASR is the core: audio-encoder + projector + LLM-decoder models trained through modality alignment, instruction tuning and RL post-training (GRPO, DPO) on clinical-entity rewards. Alongside, CTC / RNN-T / TDT models (Parakeet, Conformer families) for low-latency streaming, all on multi-thousand-hour medical and Indic corpora.
- Win on drug names and clinical entities. Extend the contextual-biasing stack (G2P phonetic matching, WFST biasing, shallow fusion with medical LMs) so that medicines, dosages and investigations are transcribed right the first time.
- Ship to production. Serve models on GPUs: quantisation, distillation, batching, latency/throughput tuning, and A/B evaluation inside EkaScribe.
- Own the data and eval engine. Build and maintain the data pipelines, test sets and evaluation framework that tell us whether a model is ready for doctors.
- Publish and open-source. Release model cards, benchmarks and papers under Apache 2.0.
- Raise the bar for the team. Review experiments, mentor engineers, and turn research findings into reusable training recipes.
Skills required
Top skills
- ASR models
- Speech-LLMs
- PyTorch
- Multi-GPU training
All skills (8)
- CTC and transducers
- Evaluation design
- WER and CER metrics
- Audio fundamentals
- Code-mixed Indic ASR
- Contextual biasing
- Model compression and serving
- Healthcare domain
Qualifications
Must have
- 2-6 years building ASR or speech models that reached real users, with at least one model you trained from data to deployment.
- Deep, hands-on fluency in one modern speech toolkit and the architectures behind it: CTC, transducers, attention encoder-decoders, SSL encoders, and speech-LLMs. You have aligned an audio encoder to an LLM, or fine-tuned one (Qwen-Audio, Whisper-encoder + LLM, Canary, Gemma-audio style), and know where they beat and where they lose to transducers.
- Strong PyTorch and multi-GPU training skills: mixed precision, distributed data loading, checkpoint surgery, and the instinct to find why a run is slow or diverging.
- Comfort with the full evaluation story: WER/CER and why they lie, entity-level metrics, calibration, and designing test sets that predict production behaviour.
- Working knowledge of audio fundamentals (resampling, codecs, VAD, noise and reverberation) and of how speech data gets messy in the field.
- Clear written communication: you can turn an experiment into a one-page finding others can act on.
Nice to have
- Code-mixed or low-resource Indic language ASR, including familiarity with Devanagari / Dravidian script transliteration.
- Contextual biasing, hotword boosting or LM fusion in production.
- RL or preference-based post-training for sequence models; hallucination control and timestamping in LLM-based ASR.
- Model compression and serving: TensorRT, ONNX, CTranslate2, Triton, vLLM-style batching.
- Publications or open-source contributions in speech; prior exposure to healthcare, call-centre or legal dictation domains.
Other details
Location: Bangalore (In-office)
Experience: 3 - 6 years
Company info
About Eka.Care
EkaCare is India's connected healthcare platform: an EMR that doctors run their practices on, a personal health record used by millions of Indians, and one of the deepest integrations with India's ABDM digital-health rails.