HackMDIBM InterConnect 2017
Disclaimer
- 4.5 days
- 2,200 session
- 25,000 people
Focussed on
- Cloud
- Watson
Neglected
- IoT
- Blockchain
- Security
Presentation agenda
- IBM vision
- IBM cloud platform
- Personal takeaways
Keynote Speech
Ginni Rometty
IBM CEO
IBM Cloud
- Enterprise-strong
- Data-first
- Cognitive to the core
Enterprise-strong
Scale (50 data centres in 19 countries)
Largest commercial IoT platform
Public, but designed for industry
Data sovereignty: choice and consistency
- On- or off-premise
- Level of privacy
- Range of services and applications
Support for hybrid solutions
- Data layer
- Application layer
Secure
Comprehensive innovation roadmap
Data-first
Value-adding strategy
- Democratisation (Facebook, Google)
- IP (IBM)
Governance
- Fine-grained access control
- Public
- Private
- Licenced
- Locality
- Isolation
Cognitive to the Core
Range of data types
- Text
- Image
- Voice
- IoT sensors
Range of capabilities
- Analytics
- Machine learning
- Cognitive computing
Prepackaged, curated domain knowledge
IBM Strategic Overview
- Platform: Watson Data Platform (WDP)
- Ecosystem: Watson Data Platform Partners
- Methodology: DataFirst Method
Watson Data Platform "Experiences"
- Data Connect - data engineers
- Watson Analytics - business analysts
- Data Science Experience (DSx) - data scientists
- Bluemix - app developers
DSx Value Proposition
- System administration
- Best practices and staying current
- Collaboration
- Sharing (internally and wider community)
- Documentation
- Governance
- Granular access control (management and security)
- Auditing
- Model deployment and integration
- Scaling
Ecosystem
Partnering with third-party vendors (mostly open source-based)
DataFirst Method
- Briefing & vision
- Discovery workshop
- Design & validate
- Implement
- Run & Maintain
Takeaways
IBM Weaknesses
- IBM is internally fragmented
- Watson's current capability ranges from weak to mediocre (evaluated against both marketing claims and competitor capabilities)
Cloud is big
API economy is big
Deep learning is big
DevOps is big
Container technology is big
Local industry lags by several years
New generation of core analytics technology is exclusively open source
IBM fully embraces and actively supports and contributes to open source
IBM explicitly supports a mix-and-match approach to using cloud services
IBM's strength is in integration and interoperability
- IBM provides tools
- Business partners
- Build solutions
- Market
- Provide support
Cloud migration is a process rather than an event
IBM advocates and technologically supports hybrid architectures (on-premises cloud)
Hybrid cloud architecture is complex and requires significant and varied skills
Many, especially smaller, businesses are confused, uncertain, and apprehensive about cloud transition
Solutions have to be designed around client needs
Biggest cloud issues for clients
- Security
- Architecture
- Migration strategy
Cloud and cognitive strategy details are very sensitive to details regarding
- Time frames
- Costing
Cloud Strategy
Liz Herbert
from Forrester
fin
