Data and Analytics Value Streams: AI-Ready Edition

The Data and Analytics Value Streams: AI-Ready Edition is for architects and transformation leads in financial services firms who need to see how data and analytics work runs from trigger to value, and where AI belongs in it. It covers 12 value streams in 60 stages. Each stage is shown as it runs today and as it could run with AI, with the target AI mode, the agents involved and the human checkpoint that stays with a person. It is the value-stream part of Business Architecture for Data and Analytics: AI Transformation Blueprint, sold on its own.

What's Inside

The same 12 value streams come in three files: a deck to present, a workbook to work in and a guide to read. The workbook holds every field; the deck shows each stage as it runs today and AI-native, with its AI mode and human checkpoint.

• 12 data and analytics value streams and their 60 stages, each with its steps, stakeholders, entry and exit criteria, and the capabilities that enable it
• For every stage: today's pain points, the AI-native future state, the agents involved, the target AI mode, the human checkpoint and KPIs
• 13 standard data and analytics value streams with their steps, as a reference process view
• The Value Streams deck (PowerPoint): every value stream as a flow, then stage by stage
• The workbook (Excel): the value streams and every stage, in sheets you can filter and extend
• The guide (Word): the scales used at each stage (AI mode, human oversight, risk tier and work pattern), every value stream in prose, and the license
• Free corrections, and new versions free for 180 days

Every file names your organization and carries your license ID.

The 12 Value Streams

Each value stream runs from a trigger to the value it delivers. The list gives that value and the number of stages.

• Business Strategy to Data and Analytics Plan (5 stages): A funded, sequenced data and analytics plan traced to the business strategy, with a target data architecture, a governance structure and named decision makers, and benefits measured against the investment cases that justified it.
• Data Need to Delivered Data (5 stages): A documented, quality-checked data set that consumers with an approved purpose can reach, built once on an onboarded source and running on schedule within its cost limits.
• Data Issue to Trusted Data (5 stages): Data corrected at its source and in its copies with the Data Owner's approval, the cause removed, and the consumers affected told the issue is closed.
• Master Data Request to Governed Record (5 stages): One trusted master record for each real entity, approved by the Data Steward, placed in its hierarchies and synchronized to every system that uses it.
• Analytics Question to Decision Insight (5 stages): An analysis a decision owner can act on, with its method reviewed, its limits stated, and the outcome of the decision tracked.
• Metric Proposal to Certified Report (5 stages): One definition and one calculation for each official metric, figures relied on externally certified before release, and reports that reach the people entitled to them.
• Access Request to Approved Data Use (5 stages): Access decided by the data's owner on its evidence, granted by cybersecurity only as approved, and use kept to the approved purpose.
• Sharing Proposal to Shared Data (5 stages): Data shared under agreed terms by an approved, secure channel, in the least identifiable form that serves the purpose, with recipients' obligations tracked.
• Data Initiative to Responsible Data Use (5 stages): A data use that has been assessed for privacy, ethical and bias risk, designed to keep to the data its purpose needs, approved with a person accountable for decisions about individuals, and checked against its purpose.
• Data Creation to Disposal (5 stages): Every data set recorded with an owner, a classification and a retention period, kept as long as it is needed and permitted, held when a hold applies, and disposed of only with the Data Owner's approval.
• Training Data Need to Delivered Training Data (5 stages): A documented training data set, assessed for fitness, representativeness and bias, with its lineage and terms of use, released to AI Management by the Data Science Lead.
• Data Product Idea to Adopted Data Product (5 stages): A data product in use, with a named Data Product Manager, published commitments and terms of use, and measured quality, freshness, availability and cost.

Who Uses It

Process owners and the teams who redesign data and analytics work, transformation leads who need a future-state process view to plan from, and business architects who map value streams to capabilities.

How It Fits With the Other Products

Each stage names the capabilities that enable it; they are defined in the Data and Analytics Capability Model: AI-Ready Edition. The agents each stage involves are specified in Business Architecture for Data and Analytics: AI Transformation Blueprint, which also holds the business data model, use cases, controls, an operating model and a roadmap. The Data and Analytics Business Data Model is sold on its own too.

Target AI modes describe a typical large enterprise and are a starting point for your own.