Senior AI Research Scientist

Location:

Working Days:

Working Hours:

Experience:

Apply:

Subject Line:

Transform+

Senior AI Research Scientist

• Formulate and drive a 12‑ to 24‑month research roadmap covering multilingual speech recognition, conversation summarisation, LLM prompt optimisation, retrieval‑augmented generation (RAG), and behavioural scoring.
• Publish internal white papers and, where strategic, peer‑reviewed papers or patents to establish technological leadership.

• Design and train advanced models (e.g., Whisper fine‑tunes, Conformer‑RNN hybrids, transformer‑based diarisation, LLM fine‑tuning with LoRA/QLoRA).
• Build rapid prototypes in PyTorch or TensorFlow; benchmark against latency, accuracy, and compute cost targets relevant to real‑time use cases.

• Work closely with backend and MLOps teams to convert research code into containerised, scalable inference micro‑services.
• Define evaluation harnesses (WER, BLEU, ROUGE, accuracy, latency) and automate regression tests before every release.

• Lead data‑curation efforts: multilingual audio corpora, domain‑specific fine‑tuning datasets, and synthetic data pipelines for low‑resource languages.
• Establish annotation guidelines, active‑learning loops, and data quality metrics.

• Act as the principal technical advisor in customer POCs involving custom language models, domain‑specific ontologies, or privacy‑sensitive deployments.
• Mentor junior researchers and collaborate with product managers on feasibility assessments and success metrics for AI‑driven features.

• 6–10 years of hands‑on research in ASR, NLP, or multimodal AI, including at least three years in a senior or lead capacity.
• Strong publication record (top conferences such as ACL, INTERSPEECH, NeurIPS, ICLR, EMNLP)

  or

  patents showing applied innovation.
• Expert‑level Python and deep‑learning fluency (PyTorch or TensorFlow); comfort with Hugging Face, OpenAI APIs, and distributed training.
• Proven experience delivering research outputs into production systems with measurable business impact.
• Solid grasp of advanced topics: sequence‑to‑sequence modelling, attention mechanisms, LLM alignment, speaker diarisation, vector search, on‑device optimisation.

• Experience with Indic or Arabic speech/NLP, code‑switching, or low‑resource language modelling.
• Familiarity with GPU orchestration, Triton inference servers, TorchServe, or ONNX runtime optimisation.
• Prior work on enterprise call‑centre datasets, sales enablement analytics, or real‑time speech pipelines.
• Doctorate (PhD) in Computer Science, Electrical Engineering, or a closely related field from a Tier 1 institution.

• Reduction of transcription error rate and/or inference latency by agreed percentage targets within 12 months.
• Successful deployment of at least two novel AI modules into production with adoption across Tier‑1 client accounts.
• Internal citation and reuse of developed components in other product lines.
• Peer‑recognised technical leadership through mentoring, documentation, and knowledge sharing.

careers@darwix.ai

Joining Darwix AI as a Senior AI Research Scientist means shaping the next generation of real‑time, multilingual conversational intelligence for enterprise revenue teams worldwide. If you are passionate about applied research that moves the business needle, we look forward to hearing from you.