Analytics Legends logoAI & Analytics Legends
Menu
Sign inSign up
Challenges

The real SAP AI & analytics problems today

As of 2026-11-06

Edited by · How this page is made

Ranked on 531 customer questions at SAP TechEd 2026, our live adverts and Search Console queries. Updated on 6 November 2026.

Map of the SAP AI and analytics landscape in five bands, people, agents, platform, data and systems, with twelve numbered challenges.
The SAP AI and analytics landscape with the twelve challenges numbered. Dashed amber means announced, rose means end of maintenance.

Start from your role

CIO or architect at an SAP customer

  1. 1 · SAP AI agents in production: identity, authorizations, governance, audit log
  2. 2 · Joule value and adoption: what shipped, what is announced
  3. 4 · AI-ready data: data products, knowledge graph, master data, custom fields

Consultant

  1. 4 · AI-ready data: data products, knowledge graph, master data, custom fields
  2. 1 · SAP AI agents in production: identity, authorizations, governance, audit log
  3. 5 · On-premise and Private Edition: what SAP AI actually reaches

Consulting firm

  1. 4 · AI-ready data: data products, knowledge graph, master data, custom fields
  2. 1 · SAP AI agents in production: identity, authorizations, governance, audit log
  3. 2 · Joule value and adoption: what shipped, what is announced

Recruiter

  1. 9 · Skills shortage: staffing an SAP AI and analytics team
  2. 12 · End of BI Platform 4.3 and the dashboards versus chat question
  3. 4 · AI-ready data: data products, knowledge graph, master data, custom fields

The twelve challenges

  1. 1SAP AI agents in production: identity, authorizations, governance, audit login phases
  2. 2Joule value and adoption: what shipped, what is announcedin phases
  3. 3Predicting SAP AI cost: AI Units, capacity units and model pricingpartly shipped
  4. 4AI-ready data: data products, knowledge graph, master data, custom fieldspartly shipped
  5. 5On-premise and Private Edition: what SAP AI actually reachespartly shipped
  6. 6Open ecosystem: Copilot, Databricks, Fabric, Snowflake, BigQuery, MCP and A2Apartly shipped
  7. 7The BW 7.5 cliff: end of maintenance 2027, extended 2030, and the migration routesend of maintenance
  8. 8BDC, Datasphere, SAP Databricks, lakehouse: which is whichpartly shipped
  9. 9Skills shortage: staffing an SAP AI and analytics teampartly shipped
  10. 10Roadmap volatility: dates that slip and how to plan around themannounced
  11. 11Compliance: the delayed EU AI Act, sovereignty and Schrems IIannounced
  12. 12End of BI Platform 4.3 and the dashboards versus chat questionend of maintenance

How the ranking works: each challenge is scored on how often it comes up, in TechEd questions, knowledge units, adverts, search queries and news, then weighted by what is at stake: money, a deadline or regulatory risk. Status words are read from our SAP announcement registry, so they move when the registry moves. A date from SAP is a plan, not a commitment.

AI agents reach the same answers through the find_answers tool of our MCP server, which returns the short answer, what shipped, what is announced and the page to cite.

Statuses change every month. Receive the monthly status update

Looking for a specific question? All guides