Founding Artificial Intelligence Engineer

Company Description

ASTA Health Tech is turning every hospital bed into an intelligent one. Our device agnostic platform uses computer vision to read any existing bedside monitor, with no replacements, no IT project, and no workflow change, transforming general wards into ICU grade smart units in under 30 minutes. We are live in multiple hospitals across India, in daily clinical use since 2025. Being live at the bedside for as long as we have, in the kinds of environments we operate in, creates opportunities that are difficult to replicate from the outside.

We are backed by MeitY Startup Hub (Govt. of India), IISER Pune, NIT Andhra Pradesh, and NIT Trichy, with scientific advisors from IIT Madras (Wadhwani School of AI) and NIT Trichy. Our founding team combines 30+ years of surgical practice, 20+ years of AI and computer vision leadership at Microsoft (M365, Azure), and deep ML expertise.

If you want to build AI that measurably keeps people alive, alongside a team that has shipped it, read on.

Why this is not a typical early-stage bet

• Product is already deployed and in daily clinical use since 2025, not idea stage.

• Real hospital customers, expanding deployments, and a clear per bed subscription revenue model.

• Sustained operation in live clinical environments, which creates conditions for building and validating systems that cannot be replicated in offline development setups.

• Clinical validation in real hospital environments, not lab benchmarks.

• Founding team that has shipped at scale before and has deep clinical and ML credibility. The problems we work on involve real world variability, noisy signals, and system constraints that do not appear in benchmark datasets. We are at the stage where core system decisions are still being defined, and early team members will shape what this becomes.

Role Description

You will design, train, and optimize computer vision models that read vital signs from bedside monitors under real world clinical conditions. You will contribute to early stage research work that sits beneath our product roadmap. You will work on problems involving real time signal extraction, robustness across heterogeneous devices, and downstream modeling on evolving clinical inputs. You will own core parts of the signal extraction and modeling pipeline end to end. The most interesting work ahead of us is in building on top of these signals, where modeling meets clinical reasoning. We will share more in the interview. You will build robust ML pipelines that run on the edge and in the cloud, with the reliability that clinical environments demand. You will collaborate directly with our CMO, a practicing surgeon, on problem framing, failure analysis, and clinical validation.

In your first 90 days, you will likely:

• Train and evaluate models across multiple monitor types.
• Improve robustness of signal extraction under real world conditions.
• Deploy and observe model behavior in live clinical settings.

Compensation and the Trade Off

This is an equity first founding role. Early team members are compensated through meaningful ownership in the company, structured to reflect founding level contribution rather than standard employee grants. We expect to introduce cash compensation post seed, targeting Q3 2026, at which point standard W 2 employment begins.

This role is best suited for individuals who would otherwise consider founding or joining a very early stage startup. You are taking asymmetric upside risk in exchange for meaningful ownership, real influence on product and company direction, and the chance to build something that measurably keeps people alive.

Eligibility

Open to US citizens and green card holders only. Fully remote within the US. Due to current payroll and compliance setup for our US entity, we are temporarily limited in the states we can hire from and are not currently able to consider candidates residing in California, Washington, Massachusetts, New Jersey, Illinois, Oregon, New York, Colorado, or the District of Columbia. We expect to expand in Q3 2026.

Qualifications

• Strong foundations in machine learning, deep learning, and CS fundamentals.

• Proficiency in Python and PyTorch or TensorFlow.

• Hands on experience with neural network architectures for vision, sequence modeling, or both.

• Experience taking a model from notebook to deployed production system.

• Master's or PhD in CS, AI, Data Science, or equivalent demonstrated work.

• Clear written and verbal communication.

Preferred

• Experience with transformer architectures or time series foundation models.

• Exposure to NLP or multimodal modeling in clinical or scientific domains.

• Prior work on physiological signals or clinical time series.

• Healthcare AI experience.

• Open-source contributions or published research.

This role is not a fit if

• You require immediate cash compensation.

• You prefer well defined scope and stable environments.

• You are optimizing for brand name experience over ownership.

Why people join us

• They wanted to work on real world clinical impact, not benchmarks.

• They wanted to build from first principles, not incrementally improve existing systems.

• They wanted ownership over systems that will scale across hospitals globally.