How Manufacturers Are Adding AI Without Replacing SAP or Oracle

Executive summary

Manufacturers have begun to augment existing enterprise resource planning (ERP) systems with artificial intelligence (AI) rather than attempting risky full–system replacements. The drivers include a wave of post‑pandemic digital‑transformation spending, increasing operational complexity, tighter inventory and working‑capital requirements, and executives’ desire for rapid return on investment (ROI).

Research indicates that by 2027 roughly 75 % of companies will shift from monolithic ERP systems to modular architectures designed to work alongside AI services. The global market for AI in ERP was valued at about US$ 4.5 billion in 2023 and is forecast to exceed US$ 46 billion by 2033 (a 26.3 % compound annual growth rate). The playbook below summarises trends, enterprise readiness, leading vendors and use cases, comparing legacy providers with innovative startups and highlighting investment patterns and case studies.

Market landscape and enterprise readiness

Adoption trends

• AI is pervasive but not yet scaled: In McKinsey’s 2025 survey, 88 % of organizations reported using AI in at least one business function, up from 78 % the previous year; however only about one‑third are scaling AI programs enterprise‑wide. The majority remain in pilot stages.
• Agentic AI emerges: Twenty‑three percent of respondents said their organisations were scaling agentic AI systems (autonomous agents) in at least one function and another 39 % were experimenting, but in any given function no more than 10 % of companies are scaling agents.
• Productivity impact: OpenAI’s 2025 enterprise AI report shows that ChatGPT Enterprise seats grew 9× year‑over‑year and that 7 million workplace seats were in use. Workers reported saving 40–60 minutes per day and 75 % said AI helped them complete tasks they previously could not.

Digital‑maturity and budgeting

• Mid‑level digital maturity: The Manufacturing Leadership Council’s 2025 survey found that 75 % of manufacturers consider themselves mid‑level in digital maturity and 80 % believe self‑managing AI‑powered facilities are coming. Only 34 % considered AI very significant—an increase from 10 % the previous year.
• Data‑quality & talent barriers: Integrate.io’s 2026 analysis reported that 64 % of organisations cited poor data quality as the top barrier to digital transformation and 95 % of IT leaders identified integration issues. Talent shortages are acute; up to 90 % of organisations face IT‑skill deficits, potentially costing US$ 5.5 trillion by 2026.
• Investment momentum: Deloitte’s 2025 survey of 600 manufacturing executives indicated that 80 % plan to invest ≥20 % of their improvement budgets in smart‑manufacturing initiatives such as automation hardware, data analytics, sensors and cloud computing. Nearly one‑quarter of manufacturers plan to adopt physical AI (humanoid or embodied robots) in the next two years.
• Digital‑transformation spending: Global digital‑transformation spending is projected to reach nearly US$ 4 trillion by 2027 with a CAGR of 16.2 %. Manufacturing is among the leading sectors in this surge.

Why augment instead of replace?

Legacy ERPs are deeply embedded in production processes, supply chains and financial systems. Full replacement projects are expensive, disrupt business continuity and frequently fail to deliver promised benefits. AI augmentation lets manufacturers build a “smart layer” on top of the existing ERP, leveraging its data while leaving core transaction processing intact. According to Gartner/IDC forecasts, by 2027 about 75 % of enterprises will add modular AI capabilities instead of replacing their ERPs. Augmented ERPs improve forecasting accuracy, reduce manual workload and accelerate decision‑making without the risk and cost of a rip‑and‑replace project. McKinsey estimates that AI‑based forecasting can reduce errors by 20–50 % and cut lost sales by up to 65 %.

Top AI‑enabled ERP platforms

The summary below encompasses ten leading AI‑enabled ERP solutions (mixing traditional vendors and newer platforms) and the AI capabilities they emphasize. The classification of each vendor as legacy (long‑established ERP providers) or startup/overlay (AI‑native or integration‑focused) highlights the diverse approaches to augmentation.

• Palantir Foundry (AI‑overlay platform): Offers an AI‑driven decision platform rather than a traditional ERP. Acts as a data integration and predictive‑analytics overlay that unifies disparate systems and provides cross‑functional intelligence. Ideal for organisations seeking to augment legacy ERP with decision intelligence. Startup/overlay: Not an ERP vendor; built as an AI decision layer. Focuses on predictive analytics, supply‑chain optimisation and simulation.

• SAP S/4HANA with Joule digital assistant: Full‑suite ERP with deep industry alignment. Embedded AI includes conversational assistants, predictive modelling and process automation. Uses a dataset from over 30 000 customers to train algorithms. Legacy: Market‑leading ERP provider modernising its core via AI and the Joule digital assistant. Investments emphasise industry‑specific AI and process automation.

• Salesforce (CRM + platform): CRM platform with AI via Einstein across sales, service and marketing. Uses machine learning for lead scoring, content recommendations and chatbots. Offers a platform for building ERP‑like capabilities on Force.com. Legacy platform gradually expanding into ERP. Strength lies in customer‑facing AI and integration with partner ecosystems.

• Microsoft Dynamics 365 with Copilot: Unified ERP/CRM solution. Copilot uses large language models to deliver real‑time assistance in finance, supply chain and sales; integrates naturally with Microsoft Office tools like Excel. Legacy: Combines ERP with productivity suite; uses OpenAI partnership. Lowers adoption barriers via familiar interfaces.

• Oracle Cloud ERP: Comprehensive cloud suite with AI across finance, supply chain, procurement and HR. Uses large industry‑specific datasets to automate financial close, optimise inventory and recommend staffing. Legacy: Deep integration across Oracle ecosystem; emphasises autonomous database and adaptive intelligence.

• Workday: Human‑capital‑management and financials platform. Built with machine‑learning models for talent acquisition, workforce analytics and financial forecasting. Legacy: Cloud‑native but established in HR/finance. AI is central to product architecture.

• Oracle NetSuite: Mid‑market cloud ERP. AI capabilities include automated financial close, supply‑chain forecasting and anomaly detection. Legacy (mid‑market): Simpler, standardized AI features; accessible to growing companies.

• ServiceNow: Platform for workflow automation and IT service management. Provides NLP‑driven case resolution, HR ticketing, predictive insights and cross‑workflow orchestration. Legacy/Platform: Strong at unifying service workflows; uses AI to deliver autonomous service management and operations.

• Rapid (now DCI/low‑code integration)**: A composable, low‑code integration platform that orchestrates microservices and unifies data across legacy systems and SaaS platforms. AI assistants manage cross‑system workflows. Startup/overlay: Focus on integration and composability rather than core ERP. Enables AI‑enabled orchestration between modules.

• Odoo: Open‑source ERP for small/mid‑sized businesses. Provides AI‑powered order processing, accounting automation and financial close. Low licence cost and high flexibility. Startup/SMB: open‑source community; accessible path to AI‑augmented ERP for SMBs.

How legacy vendors and startups approach AI augmentation

Legacy providers

Traditional ERP vendors such as SAP, Oracle, Microsoft, Workday and Salesforce integrate AI into their existing suites. Their strategies typically involve:

• Embedding AI into core processes (financial close, supply‑chain planning, workforce management). SAP’s Joule assistant offers conversational interfaces and predictive modelling; Oracle Cloud uses adaptive intelligence to automate finance and procurement.
• Leveraging proprietary datasets built over decades of enterprise implementations (e.g., SAP’s 30 000‑customer training set).
• Extending ecosystems through AI partners; Microsoft integrates OpenAI models into Dynamics 365 Copilot; Salesforce uses its Einstein platform for CRM insights.
• Providing compliance and support needed for regulated industries. These vendors guarantee data governance, security and long product support lifecycles.

Startups and overlay platforms

New entrants aim to overlay intelligence on top of incumbent ERPs rather than replace them. Common traits include:

• Composable architecture and low‑code integration allowing rapid deployment across heterogenous systems (Rapid/Decision Cloud).
• Decision intelligence and agents: Palantir’s platform unifies data and applies AI to orchestrate cross‑functional decisions. Aera Technology provides virtual AI agents that autonomously manage supply‑chain workflows and optimize decision‑making.
• Process mining and sustainability monitoring: Celonis uses process mining to find inefficiencies; Prewave’s AI radar monitors global news and social media to flag supplier sustainability risks.
• Focus on specific functions like demand forecasting, inventory optimisation or supplier risk, often delivering faster ROI.

Startups leverage cloud‑native design and advanced AI/agentic capabilities to orchestrate actions across the existing ERP and other systems, providing agility without locking customers into a single vendor.

Case studies and evidence of value

Rolls‑Royce & Microsoft (legacy augmented by AI): Rolls‑Royce partnered with Microsoft to optimise engine design and maintenance. AI models were used to design turbine blades, inspect blades and monitor engine health, resulting in a 30 % boost in machine usage, near real‑time fault resolution and detection of approximately 400 unplanned maintenance events per year. The integration kept Rolls‑Royce’s existing ERP and engineering systems intact while augmenting them with AI analysis.

Toyota & supplier risk evaluation: Toyota employs AI‑driven risk tools to evaluate suppliers’ delivery performance, consistency and quality, enhancing supplier selection and supply‑chain reliability. This overlay approach improves procurement decisions without replacing core ERP functions.

BMW – Car2X & AIQX platforms: BMW’s Car2X system turns vehicles into active supply‑chain participants. Vehicles self‑analyse and detect assembly errors, verify missing parts and communicate via sensors; the AIQX platform uses algorithms for automated quality control and defect detection. These AI overlays rely on ERP data but run separately, demonstrating how automotive manufacturers augment existing systems with AI for real‑time quality assurance.

ZF Friedrichshafen – AI Tech Center: At ZF Friedrichshafen, AI analyses customer requirements, shortens development cycles and uses virtual sensors to calculate torque, friction and service life across the value stream. The system works across the entire development pipeline, allowing earlier detection of inefficiencies and risks.

Audi & Prewave sustainability radar (startup overlay): Audi implemented an AI‑powered sustainability radar developed with Prewave to monitor supply‑chain risks across 150 countries. The system scans news and social media to detect sustainability violations or disruptions, enabling Audi to proactively take corrective action or terminate high‑risk suppliers. This overlay sits beside SAP’s ERP rather than replacing it, demonstrating targeted AI augmentation.

dotData at ADK Marketing Solutions (AI for forecasting) ADK Marketing Solutions replaced part of its TV‑audience prediction process with dotData’s automated machine‑learning system. The AI solution reduced prediction errors by 20 %, accelerated predictions by 30–40 %, and improved advertising effectiveness.

Mitsubishi Electric (Oracle Cloud AI): Implementing Oracle Cloud ERP with AI and process automation enabled Mitsubishi Electric to achieve 60 % higher uptime, 30 % increase in production, 55 % reduction in manual processes and an 85 % reduction in floor space. This demonstrates the tangible ROI of AI augmentation within a legacy ERP context.

Deloitte’s manufacturing outlook notes that US $500 billion in private sector commitments were announced by July 2025 to expand the US chipmaking ecosystem, supporting more than 500 000 jobs. The rise of AI‑driven data‑center infrastructure is creating downstream demand for industrial equipment and encourages manufacturers to invest in automation and AI capabilities.

Investment patterns and ROI considerations

• Budget allocations: Smart‑manufacturing initiatives now command a significant share of investment budgets; 80 % of executives plan to allocate at least 20 % of their improvement budgets to these initiatives. This indicates readiness for AI adoption and highlights the importance of a clear business case.
• ROI emphasis: High performers in McKinsey’s 2025 survey — only 6 % of respondents — attribute at least 5 % of enterprise EBIT to AI. Their success is linked to redesigning workflows, scaling faster and investing more aggressively.
• Data & integration costs: Poor data quality and integration challenges remain the largest obstacles; 64 % cite data quality as a top barrier and 95 % struggle with integration. Overcoming these barriers requires investment in data governance, master‑data management and open integration architectures.

Strategic recommendations for manufacturers

1. Adopt a modular AI‑augmentation strategy: Avoid rip‑and‑replace projects; instead, layer AI capabilities on existing ERP by deploying AI agents, decision‑intelligence platforms and process‑specific enhancements. Use frameworks like Palantir or Aera Technology for cross‑system decision support.
2. Prioritise high‑ROI use cases: Begin with demand forecasting, inventory optimisation, predictive maintenance and procurement guidance — areas where AI has already demonstrated 20–50 % error reduction and substantial working‑capital savings.
3. Invest in data quality & integration: Create a single source of truth through master‑data management and integration layers. Without clean, accessible data, AI projects will stall.
4. Design for human‑AI collaboration: Focus on augmenting human judgment rather than automating away expertise. Provide conversational interfaces and digital assistants (e.g., SAP Joule) to increase adoption among non‑technical staff.
5. Prepare the workforce: Address talent shortages by upskilling employees and attracting external AI talent. Deloitte recommends a “build, buy or borrow” approach to workforce planning.
6. Monitor emerging technologies: Agentic AI, embodied AI (humanoid robots) and virtual sensors are on the horizon. Nearly a quarter of manufacturers plan to deploy physical AI within two years. Prepare for integration by establishing ethical guidelines and safety protocols.

Conclusion

The transition from monolithic ERP to AI‑enhanced decision platforms is underway. Manufacturers recognise that AI can deliver faster forecasting, reduced downtime and improved working‑capital management without the upheaval of replacing their core ERP. Legacy vendors are embedding AI into their suites, while startups offer overlays that unite data and orchestrate decisions across systems. Forward‑looking manufacturers are investing significant portions of their budgets into smart‑manufacturing initiatives, but success hinges on data quality, integration and workforce readiness. By following a modular augmentation strategy, focusing on high‑ROI use cases and preparing people and processes for AI collaboration, enterprises can harness AI’s transformative potential and maintain a competitive edge in the coming decade.