Services

In addition to technology licensing and support services, Automata provides various research, development, and consulting services.  For further information, please email linguist@haleyai.com or contact Paul Haley.

Over the last decade, these include:

  • Enterprise applications delivered to and reliably hosted for our clients.
  • Product and project management and development in accordance with client team needs.
  • Production of knowledge assets and AI components for commercial clients.
  • Advanced research & development of artificial intelligence.

Research and development activities typically involve or relate to:

  • policy automation, decision management or support, and compliance; rules, “business logic”
  • natural language understanding, processing, and analysis; text analytics
  • machine learning in language, perception, medicine, and commerce
  • knowledge representation and reasoning; ontology, logic, rules; embeddings
  • deep question answering and intelligent information access; intelligent agents

Paul Haley is Automata’s principal consultant.  Paul engages employees and contractors as required or appropriate to meet requirements and improve client ROI.  Teams of two to a dozen have been assembled for projects ranging from months to years.  Paul is also available for short-term consulting, such as described further below.

Typical and recent engagements include:

  • Application development
    • clinical decision support and workflow including tele-medical
    • machine-learning/machine-vision (evidenced-based medicine)
    • non-technical “knowledge graphing” tool for personalized learning
    • device-driver up work in perception (speech, vision, pressure-sensing)
  • Decisioning and Compliance work
    • commercial licensing of the Linguist to enterprise vendors
    • consulting and development of proprietary inference technologies
    • ontology and logic generation from regulations and policies (AML, KYC)
    • advisory and information access for regulations and guidelines (tax, medical)
  • Natural Language (NLU/NLP) engagements
    • quality improvement of medical test preparation via deep parsing
    • semi-automated “knowledge graphing” for personalized learning
    • ontology production via knowledge extraction from recipe corpus
    • corpus-specific/task-targeted distributed word representations (embeddings)
    • providing “gold parses” or “tree-banking” of text (e.g., STEM textbooks, legal)
  • Machine Learning applications
    • various technologies in vision, evidence-based medicine, personalized learning
    • wide-ranging in NLU/NLP, text analytics, custom embeddings, etc.
    • prior work in life sciences, program trading & portfolio management
  • Knowledge Representation deliverables
    • large deliverable covering key aspects of large and well-known ontologies for deep QA
    • ontology of cooking extracted via NLP with temporal/causal representation details
    • ontology and topic extraction and clustering via NLP & ML for personalized learning
    • ontology and topic extraction and organization for indexing of medical knowledge
    • translation of textbook knowledge from English into technical logic syntax for deep QA

Many engagements do not fit into such typical categories, however.  For one client, Paul served as an executive in charge of business planning, fund-raising, and client engagement in life science applications of a proprietary AI service integrating many large healthcare ontologies.  For another, Paul provided management consulting services concerning intelligent agent technology and market opportunities which the client affirms had material impact.  Similarly, Paul provided extensive market research concerning commercial opportunities for certain semantic technologies which our large, sophisticated client assessed superlatively.  Finally, one very well-funded startup brought Automata in to apply NLU only to take our advice that they were not yet well-positioned to do so.  Instead, they asked us to solve what we clarified as their most strategically significant technical challenge.