We’re looking to hire a (senior) software developer (F/M) interested in Clojure(Script), machine learning, natural language processing and medicine. This project is about transforming the way decisions are made about new drugs and treatments. Currently, the safety and efficacy of treatments is assessed with Randomised Controlled Trials (clinical trials), typically these trials work by splitting a patient population in two random groups: one group is given the new treatment, the other a placebo or an old treatment. The hope is that this gives an unbiased, real world, estimate of how well a treatment works.
Unfortunately, a lot of biasses are in play, and to combat them many of these trials are conducted and published each year. To give a good overview of what works and what doesn’t, experts attempt to screen and summarise/synthesise all available evidence and trials in thorough documents called systematic reviews. These systematic reviews form the cornerstone of what is called “Evidence Based Medicine”, and inform all levels of healthcare policy. But systematic reviews are extremely labour intensive to make. It sometimes takes well over a year to search, screen, extract and summarise all relevant evidence from literature.
Our project is about applying Machine Learning and Natural Language Processing to medical literature; to help experts make systematic reviews faster and more accurately. By automatically extracting relevant entities and concepts we can help the expert make screening decisions, and by storing these extractions we hope to make better search engines across the different literature databases.
The project is part of Doctor Evidence, a company based in Santa Monica, CA. At this moment a small team is working remotely (GMT+2, EU timezone) on this project full-time as a Research & Development effort, with the intention of delivering real consumer facing projects in the near future.
As a developer, you will work in a small team that will architect and build the next-generation of data science tools for Evidence Based Medicine and Health Technology Assessment. A good fit would be a developer who can work quickly while still following best practices and who thrives on new technology, and has an avid interest in medical informatics and machine learning.
Skills & Requirements
Technical Requirements
- Willing and able to work remotely. We have a distributed team with members across the globe. Western European (GMT) timezone is preferred
- Able to work autonomously, and be self-motivating. You will be working on early-stage projects where feedback and iteration are crucial
- MSc or PhD in computer science, mathematics, artificial intelligence, life sciences, or related
- You have strong communication skills and willingness to ask questions. We do not write implementation specs for you; you'll be given a business problem to solve, and are expected to understand and explore the space of design alternatives and collaborate on a solution. Questions will occur and it is important that you'll be comfortable asking them and chasing down answers
- Familiarity with modern development workflows and release cycles, using Agile-like methodologies
Experience Requirements
- You have experience in developing and maintaining rich (Clojure, ClojureScript) applications
- If Clojure is not your primary language you are willing to learn and develop your skill set further
- Experience in building working solutions based in Natural Language Processing, Data Science and Machine Learning
Bonus
- Familiarity with Evidence Based Medicine, Systematic Review Methodology or Health Technology Assessment
- Affinity with the Medical Domain (UMLS ontologies, MEDLINE/PubMed data sets, ClinicalTrials.gov)
- Strong experience with SQL databases and data warehousing
- Familiarity with semantic web technologies like SPARQL, RDF, and Linked Data
- Affinity with statistics and related mathematical fields (meta and regression analysis, Bayesian statistics)
- Experience with (functional) logic programming
- Aside from Clojure some analytics components are developed in Java, R and Python, being able to grok what's going on there is a big plus
Our Stack
- Clojure back-end (using industry standard NLP libraries)
- PostgreSQL
- Jena/Fuseki RDF triple store
- ClojureScript front-end with Reagent/re-frame
- RabbitMQ
- Python (scikit-learn, keras, numpy/scipy) for machine learning
- Bare-metal Ubuntu servers with Docker for deployment
About Doctor Evidence
Doctor Evidence (DRE) is a software and service company that creates specialty databases of relevant evidence from published clinical studies, regulatory agency-approved drug labels, epidemiological data, and other sources. Health care manufacturers including medical imaging companies, surgical device companies, diagnostic test companies, biotech and pharmaceutical companies use clinical studies and other data sources as their primary asset to obtain regulatory and industry approval, as well as to promote their products. DRE has created state-of-the-art technologies and services that enable client companies to search for, organize, analyze, and communicate on the most relevant medical evidence.
Doctor Evidence’s development team is a group of creative, energetic people who are passionate about improving human health through the intelligent use of semantic technologies. We have used this technology to develop the industry-leading analytic platform for pathway analysis and systems biology research, and are now expanding to develop new solutions for the life sciences and healthcare markets. We are passionate about our users, our products, our technology, and our vision, and are looking for intelligent, talented and entrepreneurial developers who are eager to join our world-class team.