Michael Travers
- Email: mt -at- alum.mit.edu
- Phone: (650) 219-1876
- Personal site: Hyperphor
| |
Summary of Qualifications
- Focus on human intelligence enhancement through AI integration and creative design
- Broad experience in scientific and medical applications
- Experienced in technical leadership and management roles
Professional Experience
Hyperphor, Independent Consultant1988 - present Offering consulting services in scientific software development, knowledge management, AI applications, visualization, and human-computer interaction for a variety of clients, including Parker Institute, Collaborative Drug Discovery, Apple, Science Commons, and IS Robotics. Example of work: a
data portal for PICI-sponsored multiomic brain tumor project.
Working on platform for lab instrumentation control and data capture. Rearchitected object model and data architecture for reliability, developed internal tooling to visualize collection of workflows and resources. Tools include: TypeScript, React, GCP, Postgres, Airflow
Led development of knowledge management tools as part of the Translational Engineering team, including RawSugar, a system for storing, integrating, regularizing and semantically tagging genomic and other experimental data supplied by our remote collaboration sites; Enflame, a visual query builder for a graph knowledge base; and Alzabo, an ontology specification language and visualization framework. Supported scientists doing analysis of clinical trial data, including statistical modelling and visualization. Tools include: Clojure, ClojureScript, R, Datomic.
Nuna Health, Principal Software Engineer / Data ArchitectJan 2017 - Oct 2018 Led the development of several key data platform components, including tools for schema management and data quality monitoring. The overall business objective of these tools was to reduce complexity and increase throughput of the healthcare analytics data pipeline. Time to onboard a new client was reduced by approximately 30%. Worked with the Data Operations and Data Science teams to automate and streamline their work. Contributed to architecture of next-generation platform (APIs, reusable components, microservices, high-level pipeline definition language). Led the data quality subteam Tools include: Spark, Scala, Java, Python, Airflow, d3, Looker
Vital Labs, Principal Software Engineer / Director of EngineeringOct 2015 - Dec 2016 Led the technical development of Orchestra, a platform for management of chronic disease and personal medical experimentation. The platform included a mobile app for patients, a web-based portal for physicians and administrators, and a data analysis pipeline. Worked with clinicians and designers to continually improve the platform. Developed a workflow formalism and engine for representing medical treatment patterns. Managed operations and deployment, security, regulatory compliance (HIPAA, FDA, engineering process capture). Tools include: Clojure, ClojureScript, Datomic, React, BigQuery, HazelCast
Quixey, Principal Software Engineer / Team Lead, Knowledge Engineering GroupJuly 2013 - Sep 2015 Team lead for semantic ingestion pipeline for a large search engine. Developed ontology management tools, modules for named entity extraction, normalization, and search ranking. Lead on graph knowledge base project. Designed and prototyped a variety of research systems to illustrate semantic search concepts. Contributed to the development of query understanding and ranking algorithms. Tools include: Java, Hadoop, Pig, Spark, Luigi, RDF, graph databases, a variety of natural language libraries and resources.
Worked on small team developing a complex Rails application to support cloud-based pharmaceutical drug development research. Led development of the companies first external API, created client software for integrating the CDD product into scientific workflows. Worked closely with product management to design new features. Contributed to grant writing and business development efforts. Researched and prototyped new user interface concepts. Participated in research efforts leading to scientific publications. Tools include: Rails, Knime, Common Lisp
Research and developer for the BioCyc genomic database. Developed graph algorithms and an exploratory user interface for synthetic biology reaction pathfinding. Led the design and development of SmartTables, a web-based semantic spreadsheet for bioinformatics. Rebuilt the web platform underlying BioCyc to meet modern security standards and enable more dynamic user interfaces. Did outreach with the user community to define and refine new features. Tools include: Common Lisp, d3, JavaScript
Lead architect, designer, and product manager for the CollabRx platform, which provides knowledge-based collaborative computing infrastructure for distributed medical research and personalized genomic medicine. Conducted research into knowledge and workflow representation for biomedicine; designed and built prototypes for web-based software platform to support collaboration decision-making. Designed and developed a tool to support doctor-patient communication. Tools include: Common Lisp, R, Rails, Django, graph databases
IO Informatics, Principle Software Engineer, Knowledge Systems2006 - 2007 Designed and developed tools for integrating and visualizing biological data for metabolomics, proteomics, and related domains, using semantic web technology. Took lead role in architecture, ontology, and product and user interface design of the company's flagship product. Tools include: C#, RDF
Developed applications for pharmaceutical drug discovery (combinatorial chemistry, LIMS, data integration). Responsible for user interface design, architecture, and implementation. Other responsibilities included designing and implementing an object-relational mapping facility, and redesigning architecture for operation in a worldwide enterprise environment.
After successful acquisition of Afferent, continued at MDL and later Elsevier as Principal Software Engineer and Architect. Developed platform for other drug discovery applications including logistics, LIMS, experiment planning. Lead engineer/architect for the client-side platform and data integration strategy. Designed end-user customization framework. Developed ontology and protocols to unify cheminformatics data across applications.
Worked with the Java Tools research group designing advanced interactive programming and visualization tools; explored agent-based component architectures. Worked on XML-related application tools and contributed to schema and query standards efforts. Developed web-based knowledge management tools for corporate intranet.
Technical
Experienced in design and implementation at all levels from operating system internals to user interfaces. Languages: Java, TypeScript/JavaScript, Clojure, ClojureScript, Python, R, Ruby/Rails, Common Lisp, React, C#. Databases: MySQL, Postgres, Oracle, Elastic Search, Cassandra, Datomic, BigQuery. Deployment/Cloud: EC2, Google Cloud, AWS, Heroku, Docker. Visualization: d3.js, vega, Tableau. Big Data: Hadoop, Pig, Spark. LLMs: OpenAI, Anthropic, LangChain, DSPy.
Selected Publications
Education
Miscellaneous
Coding in Place, Stanford, volunteer section leader, 2020
Behave!,
visual programming system for the Virtual Fishtank exhibit, The Computer
Museum, Boston, 1998.