Menu Schließen

IBM Planning Analytics Meets AI: What IBM Bob changes for your TM1 system

BI2run - PA x AI

You open your budget report and see: The margin is 3 percentage points below forecast. So you start digging. Open TM1, load a view, filter variances, gather comments from three systems. An hour later, you have a theory. But not yet a solid answer.

This is exactly the process IBM wants to shorten. Since early 2026, IBM’s AI agent IBM Bob can connect directly to IBM Planning Analytics, the TM1-based FP&A system used in many German and Austrian companies for corporate planning, budgeting and forecasting.

IBM Planning Analytics, also known as TM1, is an integrated platform for financial planning & analysis (FP&A). At its core is a multidimensional in-memory OLAP cube that combines planning, forecasting and reporting functions on a shared data foundation. In Germany, Austria and Switzerland, TM1 is considered the standard for complex planning models in mid-sized and large companies.

What is IBM Bob, and what does it have to do with TM1?

IBM Bob is an AI agent from IBM based on the watsonx model. It can be connected to enterprise systems via the Model Context Protocol, or MCP for short. After that, it can be operated using natural language.

The connection to TM1 runs via the Planning Analytics Agent MCP Server. You install it directly from Bob’s MCP marketplace, store the API key from your Planning Analytics Workspace and you’re ready to go. Bob then communicates with your live TM1 environment and can query and modify cubes, dimensions, views and TI processes directly.

That sets it apart from earlier approaches. Older AI integrations in Planning Analytics displayed charts at the push of a button or answered simple data questions reactively, one answer per question. Bob understands context, carries out multi-step tasks and documents its steps.

Which capabilities does the Planning Analytics Agent actually bring?

In October 2026, IBM presented three core functions:

1. Model documentation on demand

Bob analyzes your TM1 model on its own. It captures cubes, dimensions, TI processes, views and their dependencies. The result is a structured reference document with architecture diagrams, organized by business area. Anyone who has ever taken over an existing TM1 model without up-to-date documentation will immediately understand the value of this.

2. Variance analysis via natural language

You ask: “Why is the operating margin below budget in October?” Bob pulls the relevant cubes, compares budget, actuals and forecast, identifies the biggest drivers and delivers a structured analysis with a P&L overview, variances in absolute values and percentages, and a text section on the main causes. This works directly on your production data.

3. Conversational model building

Bob can make structural changes to the model. You say: “Add a new cost center to the organization dimension as a sub-unit of Sales Germany.” Bob determines the correct parent level in the hierarchy, carries out the change, checks it and cleans up temporary artifacts. No TI code needed.

These three capabilities cover tasks that TM1 users currently do manually. Often with considerable time effort.

Is this realistic for your Planning Analytics setup?

Whether these capabilities can be used depends heavily on your model and your license. We can assess that in a short conversation.

Book a meeting →

How does the technical integration of Bob and TM1 work?

The connection runs via the Model Context Protocol, an open standard for connecting AI agents to external tools and data sources. IBM uses MCP to link Bob to Planning Analytics.

Specifically, IBM provides four MCP tools:

  • get_available_tm1_servers: Which TM1 servers are reachable?
  • get_cubes_that_may_answer_query: Which cube answers this question?
  • get_data_from_data_explorer: Data retrieval via the Data Explorer function
  • lookup_potential_members: Determine dimension elements for a query

For multi-step workflows, watsonx Orchestrate comes into play. This is IBM’s platform for agent-based process automation. It can be used to integrate approval steps into Planning Analytics workflows, for example a budget approval that a head of controlling confirms via message.

Important: The entire interaction takes place within the existing TM1 permissions. Bob cannot see or change anything that the logged-in user is not allowed to see or change.

What actually changes for TM1 admins and controlling teams?

For model developers and admins, the documentation function offers the most direct relief. TM1 models grow over the years and lose their documentation along the way. Bob closes this gap, at least for the structural view of cubes and dimensions.

For admins, the option to trigger simple dimension changes via natural language is interesting, instead of writing TI processes or navigating manually in PAW.

For end users in controlling, the entry point changes. Instead of first asking a colleague with TM1 knowledge or configuring a view themselves, they can ask questions directly.

QueBIT, a long-standing Planning Analytics partner, emphasized in its analysis: The AI integration is promising, but its benefit depends on model quality and governance. Poorly structured TM1 models don’t get better with Bob. They just become accessible faster.

BI2run - technical support

When does it make sense to use it, and when not yet?

As of October 2026, the following applies:

Available today: Planning Analytics as a Service (PAaaS) and trial environments via the free IBM trial access.

Not yet available: Planning Analytics on Cloud and Planning Analytics Local (on-premises).

This is relevant for many German companies. Anyone running Planning Analytics locally or in a hybrid architecture cannot yet use the Bob integration in production.

QueBIT gave a clear assessment of the current status: “Still early.” The capabilities are real, but not yet mature for all use cases. Especially when it comes to data changes: Without careful governance and a clear authorization concept, no AI agent should write to planning data in production.

How is IBM Planning Analytics rated in 2026?

The figures from the BARC Planning Survey 26 provide context. More than 800 users of planning software took part in the survey; the results for IBM Planning Analytics are based on 45 users:

  • 91% recommend IBM Planning Analytics. For Excel, it is 41%.
  • 100% report better integration of planning with BI and analytics.
  • 98% see more transparency and traceability in the planning process.
  • Performance satisfaction: 8.5 out of 10, industry average: 8.1.
  • 4 top rankings in the BARC Survey 2026, including Leader in the category “Integrated Products for Planning and BI & Analytics”.

These figures were collected before the AI integration was widely available. The new features are arriving on an already strong platform.

What does this mean in concrete terms for existing Planning Analytics users?

Four steps make sense:

1. Check your hosting mode. If you are still on Planning Analytics Local, you are not yet ready for the Bob integration. If you are thinking about switching to PAaaS, you now have an additional reason to do so.

2. Clean up your model before AI arrives. Bob needs consistently named cubes as a foundation. Models with logic that has grown over time benefit from a clean-up step first.

3. Define governance. Which actions may Bob perform in which environments? Which changes require human approval? These questions need to be answered before deployment.

4. Test with real questions. The IBM trial environment is freely accessible. Use it to check which of your frequent analysis tasks Bob can actually take off your hands.

The AI integration in TM1 is not a revolution. It is an addition that makes certain tasks faster. The best way to find out which ones that is for you is to test it directly.

IBM Business Partner · Planning Analytics

Is Bob already a good fit for your TM1 environment?

BI2run is an IBM Business Partner for Planning Analytics. We discuss with you whether Bob is already a fit and which steps would be needed – even if the outcome is that you should wait another release or two.

Book a meeting →

Fact box: IBM Planning Analytics and IBM Bob

14 top rankings for IBM Planning Analytics in the BARC Planning Survey 26 (800+ respondents, 20 products)
291% of PA users recommend the software. For comparison: Excel is at 41%.
3100% report better integration of planning with BI and analytics
498% see more transparency and traceability in the planning process
5Performance satisfaction IBM Planning Analytics: 8.5 out of 10 (industry average: 8.1)
6IBM Bob integrates with TM1 via the Model Context Protocol (MCP)
73 core capabilities: model documentation, variance analysis, conversational model building
8As of October 2026: IBM Bob integration only available for Planning Analytics as a Service (PAaaS)
94 MCP tools: get_available_tm1_servers, get_cubes_that_may_answer_query, get_data_from_data_explorer, lookup_potential_members

Glossary

TermDefinition
IBM Planning AnalyticsIBM’s FP&A platform, based on TM1. Combines a multidimensional OLAP cube with planning, forecasting and reporting functions. The standard for complex corporate planning in the DACH region.
TM1Technical foundation of IBM Planning Analytics. Multidimensional in-memory database for complex corporate planning.
IBM BobAI agent from IBM based on watsonx. Connects to enterprise systems via the Model Context Protocol and carries out multi-step tasks using natural language.
MCP (Model Context Protocol)Open standard for connecting AI agents to external tools and data sources. Enables structured communication between the AI model and the target system.
PAaaSPlanning Analytics as a Service. Cloud-hosted version of IBM Planning Analytics. Currently the only environment with IBM Bob integration.
Planning Analytics AgentAI assistant built directly into IBM Planning Analytics, formerly called “AI Assistant”. Analyzes TM1 data based on natural language questions within the PAW interface.
watsonx OrchestrateIBM’s platform for agent-based process automation. Coordinates multi-step AI workflows and integrates external approval steps.
FP&AFinancial Planning & Analysis. Umbrella term for budgeting, forecasting, planning and reporting in a company’s finance function.
PAWPlanning Analytics Workspace. Web-based user interface of IBM Planning Analytics for planning, reporting and administration.

FAQ: IBM Bob and IBM Planning Analytics

How does IBM Bob differ from the Planning Analytics Agent in PAW?

The Planning Analytics Agent is embedded directly in the PAW interface. IBM Bob is a standalone agent that connects to Planning Analytics from the outside via MCP. Bob can use all the capabilities of the Planning Analytics Agent, but can also integrate external tools and control more complex multi-step workflows via watsonx Orchestrate.

For which companies can the integration already be used today?

Currently only for companies using Planning Analytics as a Service (PAaaS). Companies with Planning Analytics Local or Planning Analytics on Cloud have to wait for further IBM releases. IBM has announced that it will gradually extend the integration to other deployment options.

Can Bob change data in TM1 on its own?

Yes, but only within the existing TM1 permissions of the logged-in user. Bob cannot perform any actions that the user would not be allowed to perform themselves. For production planning data, there should still be a clear governance concept that defines which changes are permitted via Bob and which require manual review.

How well structured does the TM1 model need to be for Bob to work effectively?

Bob needs consistently named cubes, dimensions and elements as a foundation. Models with logic that has grown over time, inconsistent naming or undocumented calculations deliver unreliable results even with Bob. A model review is often a sensible first step before the AI integration is used in production.

What does the IBM Bob integration cost?

The exact license costs depend on the Planning Analytics contract and watsonx usage. IBM offers free trial access for Planning Analytics and Bob so you can test the integration without any purchase obligation. For binding information, it is best to consult IBM or a certified Business Partner.

What role does BI2run play in using IBM Bob in Planning Analytics?

As an IBM Business Partner, BI2run supports Planning Analytics projects, including the evaluation and setup of AI integrations. This covers analyzing whether the existing model is AI-ready, the technical setup and the definition of governance. We’ll also tell you if the right time hasn’t come yet.

Share article:

LinkedIn
WhatsApp
Facebook
Email

More articles

Any questions? Our experts look forward to your call!