What clients bring ?
“Business Challenge”
- Clearly defined sliver of the business challenge
- Qualitative & quantitative success criteria, outcomes and measures
Key Deliverables
- Use cases, journey maps & mock-ups
- Data inventory / ontology and training schema
- Model selection and flows
- Architectural & Engineering Artifacts
Key Deliverables
- Experimentation and documented model accuracy results & explainability
- Early UI / UX Prototypes
- Draft engineering criteria for production
What clients take away
“Reusable Prototype delivering Assured Outcomes”
- User tested, production ready solution (not POC) with measurable outcomes
- Detailed programmatic plan to iteratively cover business challenge spectrum with incremental solution outputs.
Key Dependencies (Client)
- Identify stakeholders, roles and requirements
- Provide access to preferred IT infrastructure training data
- Secure Business SME, Data Scientists and IT Engineer time commitments
- Review, validate and approve work products
Key Dependencies (Client)
- Business SME, Data Scientists and IT Engineer engagement
- Review, validate and approve work products
Key Dependencies (Client)
Business SME, Data Scientists and IT Engineer engagement
Review, validate and approve work products
Key Dependencies (Client)
- Identify stakeholders, roles and requirements
- Provide access to preferred IT infrastructure training data
- Secure Business SME, Data Scientists and IT Engineer time commitments
- Review, validate and approve work products
Key Dependencies (Client)
- Business SME, Data Scientists and IT Engineer engagement
- Review, validate and approve work products
Key Dependencies (Client)
Business SME, Data Scientists and IT Engineer engagement
Review, validate and approve work products
Key Tasks (Evanke)
Engage SMEs, define user persona, draft process & journey maps
Mock-up conceptual outputs
Identify / inventory data ontology & design training schema
Research AI modeling techniques and flows
Create conceptual, logical & physical solution architectures
Provision & secure AI infrastructure, AI platforms, AI software services and users
Key Dependencies (Client)
Preprocess and prepare training data
Build appropriate AI models
Train models on a subset of data
Build microservices for powering applications
Experiment models for prediction accuracy and explainability
Develop UI / UX workflows as early prototypes
Draft engineering scalability, security, etc. for the application, models and data
Key Dependencies (Client)
Test and validate models for accuracy & explain ability.
Integrate and automate model inference in appropriate workflow / decision context (e.g., SME Human-In-The-Loop)
Change manage user feedback to improve operations
Observe, analyze and manage model operational mechanics in production (e.g., model drifts)
Capture additional use cases across business challenge spectrum and draft incremental roadmap
