Senior Machine Learning Engineer

About Inspiren 

Inspiren offers the most complete and connected ecosystem in senior living. Founded by Michael Wang, a former Green Beret turned cardiothoracic nurse, Inspiren proves that compassionate care and technology can coexist - bringing peace of mind to residents, families, and staff.

Our integrated solutions seamlessly fit into existing workflows, capturing everything happening within a community. Backed by nurse specialists and powerful analytics, we provide the data operators need to make informed clinical and operational decisions - driving efficiency, profitability, and better care outcomes.

About the Role

We're looking for a Senior ML Engineer to build intelligent, scalable workflows powered by cloud-based Vision-Language Models (VLMs). You'll own the end-to-end pipeline: from determining what signals and context need to come off the device to maximize VLM performance, to building and evaluating the cloud-side inference systems that generate high-quality labels at scale. This role sits at the intersection of computer vision, LLMs, and systems engineering, and you'll work closely with our Embedded Systems, Computer Vision, and Data Science teams to close the loop between edge devices and cloud intelligence.

What you'll own

• Design, build, and iterate on VLM-based pipelines that generate high-quality labels and annotations at scale, including prompt engineering, fine-tuning, and evaluation


• Determine which signals, frames, metadata, and contextual features should be sent from edge devices to improve VLM accuracy and reduce ambiguity


• Collaborate with Embedded Systems and Hardware teams to define device-side preprocessing and data-forwarding strategies that balance bandwidth, latency, and model performance


• Collaborate with Data Science to build robust evaluation frameworks to measure label quality, model accuracy, and regression detection


• Benchmark and integrate commercial and open-source VLMs, staying current on the fast-moving landscape of vision-language capabilities

What you bring

• 5+ years of experience in machine learning engineering, with hands-on work in computer vision and/or LLM/VLM systems


• Strong familiarity with Vision-Language Models — you've used, evaluated, or fine-tuned models like GPT-4V, Claude's vision capabilities, Gemini, LLaVA, or similar


• Experience building and evaluating labeling systems at scale


• Solid understanding of how edge/device constraints (bandwidth, compute, power) shape what data is available to cloud-side models


• Proficiency with a modern ML stack: Python, PyTorch, cloud inference APIs, and tools for experiment tracking and evaluation


• Practical prompt engineering skills — you know how to get the most out of large models through structured prompting, few-shot examples, and iterative refinement


• A pragmatic engineering mindset — you care about building systems that work reliably in production, not just in notebooks


• Comfortable scoping and driving work independently in a fast-moving, early-stage environment


• Strong communication skills and a collaborative approach to working across hardware, embedded, and ML teams

Bonus skills

• Experience building model-based automation for data labeling


• Experience building infrastructure for cloud-based inference at scale, for example batching, orchestration, cost management, and monitoring


• Familiarity with edge computing, on-device inference, or embedded ML pipelines


• Experience with video or spatiotemporal data (event streams, multi-frame reasoning, temporal context)


• Healthcare or senior living domain experience

Details

• The annual salary for this role is $200,000-$230,000 + equity + benefits (including medical, dental, and vision) 


• Flexible PTO


• Location: Remote, US or Canada. NYC preferred.


• Qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender perception or identity, national origin, age, marital status, protected veteran status, or disability status.