Tallis Commercialisation

1
Tallis Commercialisation

• Tony Rose
• ACL

2
Current Situation

• Long history of previous commercialisation
  activity
• InferMed was / is main outcome
• We should aim to learn from the past
• BPM is gaining momentum
• Emerging standards (e.g. BPMN, BPEL, UML)
• Maturity investment
• Critical mass?

3
Strengths

• Key USPs
• Principled representation of declarative K
• Supports reasoning over inconsistent data
• Intuitive representation of process K
• Academic credibility
• Published spec for Proforma
• Strong brand
• Particularly among biomedical community

4
Weaknesses

• Limited interoperability with key standards
• SNOMED, XML, BPMN, HL7, etc.
• Academic culture
• Compare with Map of Medicine
• Low-key science but high-key spin
• Tallis is relatively unproven
• No real user base no user feedback
• Weak expression language
• No facility for plug play

5
Opportunities

• Open Source
• Leverage user community
• Close the feedback loop
• Synergy btw OSS commercial
• ArgoUML Poseidon
• Semantic Web
• A bandwagon, but better to be on than off it
• OWL, RDF, etc.

6
Threats

• Open Source
• This will happen anyway, with or without us
• e.g. EGADSS
• InferMed
• May decide to take their bat ball home
• The past
• Have we learnt anything? If so, what?
• Whats different this time around?

7
Options (1)

• Get Tallis out there. Now.
• Promote it (subtlely)
• Learn and improve it
• Explore OSS to
• Build a community
• Build momentum
• Engage with the BPM community
• BPM will happen, with or without us
• position ourselves as value-adders, not also-rans

8
Options (2)

• Establish a differentiated value chain, e.g.
• Make Composer OSS
• facilitates development of functionality
• establishes platform for content creation
• Promote freely available repository access
• users need access to enactment services
• help to establish critical mass of content
• Keep Tallis Engine (Tester?) proprietary
• point-of-care enactment is key link in value
  chain
• but remember ANYONE can replace it, at any time

9
Options (3)

• Technology vs. Content
• Default is to focus on technology, but
• Internet content is King (c.f. BMJ lt-gt InferMed)
• Content publishing offers great potential, but
• How to get critical mass of content/users?
• What role for open access publishing?
• Who could/would partner us?

10
Wild Cards?

• Speech interaction thru HOMEY
• Still big business
• Clear benefit to speech UI developers
• Targeted, focussed niche
• Tallis as TroubleShooter builder
• Widely used throughout s/w industry
• Facilitate development AND add smarts

11
Future Work

• Content publishing
• Explore the dynamics
• Identify value chains
• Find the niche(s)
• Need to balance content technology
• The ideal balance is probably quite subtle
• Will need to iterate several times