The posting
Why Join Sensi.AI
Sensi is redefining how care organizations support aging at home by using it’s Care Intelligence platform to translate everyday signals into meaningful clinical and operational outcomes.
Sensi’s platform is deployed across 50 states in leading home-care providers, supporting tens of thousands of seniors aging safely at home. By transforming real-world activity in the home into actionable insights, Sensi enables care teams to detect changes in condition earlier, intervene proactively, and deliver higher-quality care at scale.
This is a hybrid role based in Tel- Aviv, Israel.
About the role
As a Senior Data Scientist, you’ll use LLMs to turn the outputs of our existing models into richer, more actionable insights. You’ll own the work end to end, from defining the problem to evaluating models and shipping to production, and you’ll shape how we measure LLM quality across the team.
What you’ll do
- Own the LLM lifecycle end to end, from problem framing and model selection through evaluation, deployment, and production monitoring
- Lead our LLM evaluation: design benchmarks (synthetic and real-world) that measure how commercial and open-source models perform on our actual use cases, and define what “good enough to ship” means
- Design and build LLM pipelines that turn the outputs of our audio and NLP models into richer, clinically meaningful insights, with a deep focus on prompt engineering and RAG
- Run experiments and measure their impact in real-world systems, iterating based on data
- Processing, cleaning, and preparing data from various sources, including unstructured data
- Partner with product, engineering, and our clinical team to turn real-world needs into AI solutions, and communicate results to technical and non-technical stakeholders
- Track the state of the art in NLP, LLMs, and agentic AI, and bring what works into our products
Requirements
- 8+ years of hands-on experience in data science, ML engineering, or applied AI roles
- Hands-on experience building LLM-based solutions in production, including prompt engineering and RAG architectures
- Proven experience designing LLM evaluations: building benchmarks and eval suites, and using them to compare models and guide decisions
- Strong Python skills and experience building data pipelines for structured and unstructured data (text, and ideally audio-derived data)
- A data-driven mindset: you define success metrics up front and iterate based on results
- Self-directed, comfortable defining the problem as much as solving it
- Strong communication skills, with the ability to explain complex results to non-technical stakeholders
Nice to Have
- Experience with time series data and forecasting
- Experience with cloud platforms (AWS preferred)
- Experience with fine-tuning LLMs
- Background in healthcare, clinical, or audio-related domains
