Head of AI and NLP - Abstractive HealthCompany Background: Abstractive Health helps doctors retrieve external medical records and creates an AI automated summary using generative AI. Doctors are overwhelmed by documentation and have just 3 minutes on average to review the patient chart. We solve this problem by making the cliff notes for healthcare so doctors can focus more of their time on patient care. Our company was borne through our research with Weill Cornell Medicine where we demonstrated the capability a narrative computer-generated summary would provide for physicians. We are a VC funded startup with customers and launching a pilot at NewYork Presbyterian Emergency Department.Start Date: Immediately. You will be expected to start on-site within 3 weeks of being offered the role.Salary Range: $170,000 - $200,000 with stock options equity.Location: In-person role. 335 Madison Ave, New York, NYDescription: As a team member in NLP, you will be responsible for leading, managing and building upon our transformer models for generative AI tasks in healthcare; with particular emphasis on clinical summarization. You will work directly with the CEO on the following:
- Implementing and developing novel machine learning approaches for advancing the usefulness of our clinical summarization pipeline
- Deploying our LLMs into a cloud computing system, such as AWS
- Designing strategies to improve the speed and accuracy of our summarization and generative AI workflows.
- Data cleansing efforts with our partner medical centers for fine tuning our models
- Working with our clinical team to continually design a summary that is useful
- Writing and publishing our research in the space of clinical summarization
- Leading external research projects with health systems within the space of automating clinical documentation
- Occasionally role-switch, as needed, to be a mid-level back-end engineer role based on the startups business needs
We are currently integrated with 10 EHR systems and the national health information exchange, to pull the complete longitudinal prior medical history of the patient. You will be expected to handle a variety of unexpected data formats including CDA and FHIR. A patient record has on average 100,000+ words; we have proprietary methods for summarizing all of these words. You will work alongside us to improve this method for healthcare.You will work with our current engineering team to train, and manage them in coordination with the CEO.Benefits: We are one of the most innovative companies for generative AI and healthcare. You will be building and designing technology that will directly improve patient care and our work has a measurable impact on physician satisfaction.We offer stock options, unlimited vacation policy, paid holidays, health insurance, unlimited New York City MetroCard, casual dress, job training and conferences, and flexible work schedule.Company Culture: We are an early stage startup company with a close-knit environment among employees with strong collaboration. We work in-person everyday in New York City near Grand Central. We often pitch in and assist each other as needed. You should be a self starter, and very comfortable with problem solving and seeking help. Employees that excel are great at communicating to others on what tasks they can do well, what tasks they could comfortably learn to do quickly, and which tasks are too challenging for them to do.Requirements:
- PhD in Computer Science, Information Sciences, or similar degree program
- 2 + years of industry experience implementing and modifying large language models
- Authorship on various research papers in clinical informatics or machine learning at top conferences or journals
- Experience replicating scientific research papers in the field of NLP and/or clinical informatics.
- Demonstrate an ability to lead 9+ month long research projects from inception to completion with minimal guidance. Ensuring that the research project is consistently in line with our business objectives and stays on budget and on track.
- Strong communication skills including an ability to present research at a conference, sit on industry panels, and can communicate with CEOs and CMOs of health systems.
- Ability to mentor software and NLP engineers
- Very strong familiarity with Python, SQL, PyTorch (or Tensorflow), Pandas, Numpy, Sklearn, and Hugging Face
- Expert of the full machine learning pipeline, with significant experience, including data acquisition, data cleansing, model selection, model training/fine-tuning, hyperparameter tuning, and model deployment.
- Familiarity with deployment of LLMs on a cloud computing environment such as AWS
- Able to travel to anywhere in the world at least 3 times a year to industry events and conferences.
DEI Commitment: We are committed to fostering a diverse, equitable, and inclusive (DEI) work environment where every individual's unique background, perspective, and experience are valued and respected. We believe that DEI not only enriches our team but also drives innovation, creativity, and success in achieving our mission to transform healthcare through cutting-edge technology.We actively seek and encourage candidates from diverse backgrounds, cultures, and experiences. We strive to create a workplace that promotes equal opportunity, where everyone feels supported, empowered, and inspired to contribute their unique skills and perspectives. We are dedicated to building a team that reflects the rich diversity of the communities we serve and to making a lasting, positive impact on the healthcare ecosystem by championing DEI in everything we do.