Why Software Today Feels Like The Auto Industry Be...

In my role as a Platform Product Manager at SAP, I see every day how the software industry is entering a phase that strongly resembles the moment when the automotive industry shifted from combustion engines to electric and software-defined vehicles. On the surface, we see incremental improvements. In reality, a fundamental platform shift is happening…

….and it is leading us towards something even bigger: The Autonomous Enterprise.

What the “skateboard” platform and software-defined vehicle were for the auto industry is, in many ways, what AI-first, platform-centric architectures are becoming for enterprise software. At SAP, this is shaping our business and platform strategy. From Business AI and Joule to the SAP Business Technology Platform (SAP BTP) as the backbone of intelligent, end-to-end business processes and, ultimately, more autonomous business operations.

From Feature Software to AI-First Business Platforms

In the automotive industry, the real game changer was not just the battery. It was the completely rethought vehicle platform: centralized computing, over-the-air updates, and the idea that value creation truly starts after the car leaves the factory.

Enterprise software is undergoing a similar transition. We are moving from isolated feature development in individual applications to building platforms where data, processes and AI are tightly integrated. SAP BTP plays a central role as the data, integration and AI layer across core processes such as finance, supply chain and HR. AI is no longer “just another module”, but an integral part of how we design the platform: from embedded Business AI and Joule to AI agents that orchestrate end-to-end business scenarios and automate decisions.

Organizations that simply “bolt on” an AI feature to existing products are effectively doing what some OEMs did when they tried to electrify a combustion chassis: it works at first, but it does not scale and it does not fully exploit the potential of a new platform.

From Software-Defined Vehicles to the Autonomous Enterprise

If electrification and software-defined vehicles were the first big step, autonomous driving is the next logical stage in the automotive transformation. It builds on the same platform foundation – sensors, connectivity, centralized compute – but moves the value proposition from “assisted driver” to “self-driving system”.

The same pattern is now emerging in the enterprise:

AI first made individual processes “assisted”, through recommendations, anomaly detection, and copilots embedded in workflows.
The next step is autonomous processes that can sense, decide and act across systems: think of an order that triggers fulfillment, adjusts capacity, renegotiates freight and updates the forecast without human intervention. With humans supervising and steering, not micro-managing.
The long-term vision is an autonomous enterprise: a company that continuously optimizes itself based on real-time data, policies and goals, orchestrated by AI and agents running on a robust platform.

Just as autonomous driving requires a software-defined vehicle as its base, the autonomous enterprise requires a software-defined business: clean data, consistent process models and a platform that can host intelligent agents and automation at scale.

Generated-Image (12).Png

The Shift in Value Creation – From Hardware to Services, From Apps to Platforms, From Automation to Autonomy

In the automotive world, value creation is shifting from one-time hardware sales to software functions, data-driven services and mobility ecosystems.

We are seeing a similar pattern in enterprise software, with an additional twist:

From license-centric thinking to cloud subscriptions and platform consumption.
From isolated IT projects to AI-enabled, end-to-end business processes on a unified platform.
From task-level automation to goal-driven autonomy, where systems can propose and execute actions in line with business objectives and guardrails.

Here, SAP has a structural advantage: we are deeply embedded in mission-critical business processes around the globe and have access to highly relevant, trusted business data. Exactly the context that enterprise AI and autonomous scenarios need to be meaningful rather than experimental.

Platform Thinking from a Platform Product Manager’s Perspective

As a Platform Product Manager, I work right at this intersection. We no longer look at integration, APIs, events, data and AI services as a loose toolbox, but as a coherent business platform on which customers and partners can build their own innovation….and progressively more autonomous capabilities.

In practice, this means:

Designing and Evolving SAP BTP as an open, enterprise-grade foundation for AI and agentic use cases, not a closed island. Agents need a consistent way to access data, call business APIs and trigger workflows across systems.
Treating integration and data quality as the backbone : without a clean integration and data layer, AI remains a demo and autonomy remains a slideware vision.
Creating a developer and platform experience where APIs, events, AI services, rules and extensibility tell a single story: sensing events, reasoning over data and acting through workflows and business objects.

In other words: if software-defined vehicles needed a “central compute brain”, the autonomous enterprise needs a business platform brain – and that is where the platform strategy becomes truly differentiating.

To make this more tangible, consider the Order‑to‑Cash (O2C) process as the enterprise equivalent of the path from driver assistance to autonomous driving.

A concrete autonomous enterprise scenario – Order‑to‑Cash as the “autonomous driving” lane : In a first stage, AI improves individual steps: classifying incoming orders, predicting demand, suggesting risk scores, or flagging payment issues. This is similar to Level‑1/2 driver assistance in a car: helpful, but the driver (the user) is still in full control.

With agentic AI on/with SAP BTP and Joule, the pattern shifts: AI agents can monitor the entire O2C flow end‑to‑end, reason over real‑time business data and take actions across systems. For example:

An Order Reliability Agent continuously scans open sales orders, checks supply, lead times, logistics constraints and predicts which orders are at risk.
If a risk is detected, the agent automatically proposes mitigation options like alternative plants, substitute materials, different carriers or adjusted delivery dates, based on company policies and service levels.
A Cash Management or Disputes Agent can, in parallel, reconcile incoming payments, detect anomalies and automatically trigger follow‑ups or dispute workflows, reducing manual reconciliation effort dramatically.news.

From a platform perspective, SAP BTP provides the build, integration, data, workflow and AI services that these agents rely on: APIs to access S/4HANA or CX systems, eventing for business changes, AI engines for prediction and workflow/rules for execution. Joule then acts as the orchestrator and interface: business users can ask natural‑language questions (“Which high‑value orders are at risk this week?”) and approve, override or refine the actions that agents proposes.

This is very close to how autonomous driving stacks are built:

Sensors and maps in a car correspond to operational and planning data across ERP, CX and supply chain.
The driving policy engine corresponds to business rules, guardrails and objectives (service levels, margin, risk).
The autonomous driving controller corresponds to AI agents on SAP BTP and Joule, which continuously sense, decide and act across multiple “lanes” of the enterprise.

The endpoint is not a “hands‑off” company, but an autonomous enterprise where a large share of O2C, source‑to‑pay and plan‑to‑produce runs system‑driven. And people focus on exceptions, design and strategy. Just like a driver in an autonomous vehicle, humans stay in charge of the destination and constraints, while the platform handles the repetitive steering in between.

Legacy Thinking vs. AI- and Autonomy-First Architecture

The key leadership question, in my view, is not: “Where can we quickly add some AI?”

Instead, it is: “What would our product and platform architecture look like if we rebuilt it tomorrow, AI- and autonomy-first, from a clean slate…using the business processes and data we already have?”

A few principles I consider critical, for us and for our customers:

Architecture before features: Rethink the platform model first. Data, events, policies and agents…then add capabilities on top.
Data and model strategy at the core: Without a clear data foundation, governance and feedback loops, AI and autonomy will stay isolated pilot projects.
Platform and ecosystem mindset: Openness, APIs and co-innovation with partners and developers are the lever for scale. No single vendor will “own” all agents or all models.
AI embedded, not bolted on: Design AI and agents as a natural layer across processes, data and user journeys, with humans in control and clear guardrails.
From workflows to policies and goals: Move from hard-coded workflows towards systems that can optimize against business goals within defined policies. Similar to how autonomous driving optimizes route, speed and safety within traffic rules.

A Question for Platform Leaders on the Road to the Autonomous Enterprise

The automotive industry has learned the hard way what happens when a platform shift is underestimated. In enterprise software, we have the opportunity to avoid repeating that mistake and to use the current AI wave as a stepping stone towards the autonomous enterprise.

If you are responsible for platforms or products…at SAP, on the customer side or in the partner ecosystem, ask yourself two simple questions:

Which parts of your platform would you not build the same way again if you had to start tomorrow, AI-first, from scratch?
What would you need to change so that your systems can not only automate tasks, but autonomously optimize outcomes under clear business guardrails?

That is where your real transformation roadmap begins and the journey we will continue at SAP Sapphire this year, where the next chapter of SAP BTP’s evolution, from intelligent platform to true autonomous enterprise backbone, will come into focus.

Curious to hear:
What would you change in your platform today if your goal was not just automation, but a truly autonomous enterprise tomorrow?

SAP Business Technology Platform, Joule Joule Studio #AutonomousEnterprise SAP Sapphire SAP Business AI

Source link