Artificial intelligence — and now generative AI — is redefining how banks approach forecasting, reporting, and decision-making. For finance teams, this shift is no longer theoretical.
Leading institutions already use AI to build smarter models, improve forecast accuracy, and reduce the manual effort behind insights and reporting. The opportunity now is to operationalise these capabilities — embedding them directly into planning and performance processes.
Oracle’s Intelligent Performance Management (IPM) suite, part of Oracle EPM Cloud, introduces a powerful set of AI-driven features designed specifically for finance. It gives teams the ability to automate forecasts, generate variance narratives, and run more adaptive, scenario-based planning cycles.
In this (recently updated) blog, we explore the key IPM capabilities and how they can help finance functions deliver faster insights, greater efficiency, and more strategic impact.
Disclosure: This blog was created based on a conversation (or series of prompts) with Revvy, a ChatGPT developed by Revvence and trained on finance transformation and systems change-related content.
The vision for AI in Oracle EPM Cloud focuses on complete integration, ensuring that AI capabilities enhance every product feature. This enables the seamless use of advanced AI capabilities, including predictive analytics, AI-assisted analysis, and GenAI across all Oracle EPM applications.
.png?width=654&height=364&name=napkin-selection%20(2).png)
Oracle's approach prioritises a user-centric design. It aims to create AI tools that are not only user-friendly but also accessible to finance professionals without a data science background. The philosophy behind this design focuses on simplifying the use of advanced analytics powered by Machine Learning, making it easier for finance to fully utilise AI's potential.
Now, let’s explore the specific IPM capabilities.
Oracle’s predictive capabilities deliver powerful new capabilities for AI-powered financial planning, modelling and analysis in the finance function. These capabilities enable banks to create more advanced financial models and forecast future performance faster and with greater precision. The result is scientific-grade modelling and forecasting and substantially more agility.
As AI continues to evolve rapidly, banks that have already laid the groundwork during the digital transformation era are uniquely positioned to benefit from the speed of AI transformation. With Oracle’s powerful AI tools, financial institutions can quickly scale AI across their operations, revolutionising everything from customer service to regulatory compliance, optimising RWAs, improving product margins and automating vast amounts of reporting to position themselves as leaders in the future of banking. You can learn more about Oracle's broader AI strategy and their unique advantages in this blog, "Oracle AI Solutions for Modern Banking Operations."
Oracle’s predictive capabilities within its Intelligent Performance Management (IPM) framework are designed to transform how banks approach financial planning, modelling, forecasting, and decision-making. These capabilities are underpinned by advanced algorithms and tools that enhance predictive accuracy, streamline model-building processes, and ultimately drive better business outcomes.
The core of Oracle's predictive capabilities lies in the Predictive Planning and AutoPredict features. Predictive Planning enables users to interactively create, adjust, and validate data sets. This feature allows finance teams to engage with their data directly, making real-time modifications to assumptions and parameters and immediately seeing the results of changes.
With Oracle IPM tools, users can quickly create complex models for various use cases such as forecasting, profitability analysis, risk weighting, cost optimisation, and rolling forecasts.
The capability to conduct hundreds of what-if analyses enables banks to simulate a wide range of scenarios, leading to better-informed decision-making and strategy development. This potential is particularly valuable in areas such as strategy, group planning, stress testing, and treasury functions within banks.
A variety of sophisticated algorithms bolsters Oracle’s predictive capabilities. They now have thirteen different predictive algorithms, including:
These algorithms provide a powerful set of data science tools specifically designed for individuals who may not have a background in data science. These tools enable banks to conduct comprehensive financial modelling, including predicting the effects of changes to MEVs and evaluating the probability of loan defaults. The flexibility to select and adjust algorithms means banks can adapt their models to a wide range of operational scenarios.
Banks, especially larger institutions, often face challenges with imperfect datasets. Missing or inconsistent data can hinder accurate financial analysis and decision-making, leading to project delays as technology teams try to create a perfect dataset, sometimes at a significant cost with little to show for the investment.
The problem is exacerbated by the substantial volumes of data generated daily from various departments and systems. Oracle's predictive algorithms, such as Naive Forecaster, XGBoost, and Prophet, offer solutions to address these data quality issues. By employing advanced statistical methods and machine learning techniques, these algorithms can efficiently fill in missing values, analyse intricate relationships, and leverage historical patterns.
This capability enhances the accuracy of models and forecasts, allowing important projects to progress without waiting for the perfect data set.
AutoPredict is an advanced feature in the IPM suite that improves finance teams' modelling and forecasting capabilities. This feature uses sophisticated machine learning algorithms and statistical techniques to automate the predictive modelling process, making it simpler and faster than traditional methods.
Here’s a detailed look at what AutoPredict does and how it could help finance teams in banks.
Automated Algorithm Selection:
AutoPredict automatically selects the most suitable predictive algorithms based on the available historical data. It evaluates various algorithms, such as Naive Forecaster, XGBoost, and SARIMAX, to determine which will yield the most accurate predictions for a given dataset. This automatic selection process is guided by performance metrics, such as Root Mean Squared Error (RMSE), ensuring that the chosen model aligns with the data's specific characteristics.
Forecast vs. Prediction Variance Analysis:
Finance teams are constantly analysing variance reports to understand why actuals differ from forecasts. AutoPredict enables a more scientific analysis of the differences between forecasts and actual predictions. This feature presents finance teams with discrepancies and potential root cause analysis.
By using advanced techniques to understand these variances, finance teams can identify areas for improvement in their modelling and forecasting processes and refine their assumptions, leading to more reliable financial planning.
User-Friendly Interface:
The interface for AutoPredict is designed to be intuitive, allowing users—regardless of their technical expertise—to generate predictions quickly. The features are easily accessible across the EPM platform and can be used when finance teams are working with data in Excel. This ease of use means finance teams can leverage advanced analytics without extensive data science training.
What-If Analysis:
AutoPredict enables scenario planning through what-if analysis, allowing users to modify input variables and instantly view the impact on their models. This capability is essential for banks that need to assess the potential effects of changes such as interest rates, economic conditions, or regulatory environments on their financial models.
Bias Reduction:
One of AutoPredict's key advantages is its ability to help reduce the biases that often accompany manual forecasting. By relying on data-driven methods rather than subjective judgment, AutoPredict delivers a more objective approach to financial forecasting.
Key Takeaway
AutoPredict is a powerful feature that equips finance teams in banks with the tools needed to navigate the complexities of financial modelling at scale. By automating algorithm selection, creating variance analysis, and supporting scenario planning, AutoPredict enhances the accuracy and efficiency of financial predictions, ultimately empowering users to make better decisions faster.
Multivariate predictions are a powerful analytical capability that will soon be available in the Oracle IPM suite. It allows finance teams to forecast future outcomes by analysing multiple variables simultaneously. Unlike univariate predictions, which focus on a single variable, multivariate predictions consider the relationships and interactions between several factors, providing a more comprehensive view of potential outcomes.
Here’s a detailed exploration of what multivariate predictions do and how they benefit finance teams in banks.
Comprehensive Data Analysis:
Multivariate predictions can leverage various data sources and features, such as historical financial results, economic indicators, customer behaviours, and market conditions. This approach allows a more nuanced understanding of how different factors impact financial performance.
For example, a bank might analyse the interplay between interest, unemployment, and consumer confidence when predicting loan default rates.
Modelling Relationships Between Variables:
The algorithms used for multivariate predictions, such as regression models, decision trees, or machine learning techniques like XGBoost and LightGBM, can identify complex relationships between input variables.
These models can uncover how changes in one variable (e.g., a rise in interest rates) can affect other variables (e.g., loan demand and default rates), leading to more accurate forecasts.
Data-Driven Insights:
By analysing how various factors work together, multivariate predictions provide insights that can affirm or influence decisions. Finance teams can use these insights to adjust their assumptions, policies and strategies to optimise their product offerings and enhance risk management strategies.
Scenario Simulation:
Multivariate predictions allow finance teams to conduct what-if analyses by manipulating multiple input variables to see how they affect outcomes. This capability is crucial for scenario planning, enabling banks to evaluate potential future states based on varying economic and market conditions.
For instance, a bank might simulate the effects of an economic downturn on its product portfolio by varying interest rates, unemployment rates, and consumer spending patterns.
For example, when predicting default rates, incorporating economic indicators such as GDP growth, inflation rates, and credit scores can lead to more accurate assessments than relying on a single metric.
For example, a bank can create more accurate risk models or adjust its risk assumptions based on a better understanding of its portfolio data, leading to improved capital allocations.
Understanding how various factors influence and interact with each other means that finance teams can better align their strategies with a more scientific approach.
Integrating these predictive capabilities into the broader EPM suite further enhances efficiency, as finance teams can seamlessly incorporate insights into their planning processes.
For example, in response to a sudden economic downturn, such as a shock budget or a geopolitical event, a bank can quickly reassess the impact and adjust its strategies to mitigate potential losses or take advantage of new opportunities.
Key Takeaway
Multivariate predictions are a transformative capability within Oracle’s EPM suite that equips bank finance teams with the tools to make scientific-grade, data-driven decisions. By analysing multiple variables and understanding their interrelationships, banks can enhance the accuracy of their modelling of their forecasts, improve risk management practices, and develop strategic initiatives that will enhance financial performance.
As financial environments become increasingly complex, leveraging multivariate predictions will be essential for leadership teams to chart a sustainable growth path.
AI-Assisted Insights is a key feature within the IPM suite that uses artificial intelligence and machine learning to enhance data analysis and decision-making processes. This functionality provides automated, actionable insights that help banks navigate complex financial data, saving time and enabling more informed business decisions.
Here is a detailed exploration of what AI-assisted insights do and the possible benefits.
AI-Assisted Insights can automatically identify anomalies in financial data, such as unexpected spikes or drops in transaction volumes, discrepancies in forecasts, or deviations in expense patterns. This feature continuously monitors data streams and alerts finance teams to unusual patterns that may indicate operational inefficiencies, fraud, or compliance issues.
The tool conducts a thorough analysis of variations by comparing actual financial performance to forecasts or historical data. It quantifies differences and provides context to understand the underlying reasons for these variations.
For example, if a bank’s loan defaults rise unexpectedly, AI-Assisted Insights can highlight this variance and suggest potential causes based on historical trends and external factors.
AI-Assisted Insights employs predictive modelling techniques to forecast future financial outcomes based on historical data. This enables finance teams to anticipate trends, identify risks, and make proactive adjustments to strategies.
The tool can predict factors such as product demand, net-interest income fluctuations, and operational costs by analysing historical performance data, helping banks plan more effectively.
This feature analyses the relationships between different financial variables to identify correlations that may not be immediately evident. For example, by understanding these relationships, finance teams can make more informed and timely decisions about resource allocation and risk assumptions.
A bank looking for growth opportunities may see that a specific type of marketing spending correlates with higher loan applications, allowing it to optimise its marketing strategies based on more scientific insights.
AI-Assisted Insights can use NLP to auto-generate user-friendly summaries of complex data analyses, making insights accessible to non-technical stakeholders.
This means finance teams will spend less time creating narratives from scratch and more on activities that advance the bank's strategic objectives.
Generative AI’s potential for transformation in banking spans many areas. We've written extensively on this topic, and you can read some of the blog content specific to AI here.
Generative AI (GenAI) capabilities in Oracle’s IPM suite significantly enhance how finance teams interact with their data, mainly through three key functionalities: generating narrative summaries, asking questions about data and processes, and producing natural language charts and graphs.
Here’s an exploration of these areas and how they benefit finance teams.
Generative AI in Oracle IPM can automatically create narrative summaries that explain key findings from complex financial data. This feature translates detailed data analyses into clear and concise explanations, making it easier for stakeholders to understand important insights. It also significantly reduces the time finance teams spend on creating reports.
For instance, after analysing product performance data, the AI might generate a summary highlighting significant trends, variances, and potential areas of concern, such as increasing default rates in specific customer segments.
By automating this process, finance teams save considerable time that would otherwise be spent manually compiling reports, allowing them to focus on strategic decision-making rather than data interpretation.
With the integration of Generative AI, finance teams can engage in conversational queries about their data and processes. Users can ask specific questions like, “What would be the impact if we increased margins by .5%?” or “How do interest rate changes impact our income forecasts?”
The AI processes these inquiries in natural language, as you see with tools such as ChatGPT, and retrieves relevant insights based on the bank's data. This interactive capability means finance teams can "chat" to explore data, enabling them to analyse data more intuitively. This allows them to get the insights they want faster and without extensive training in data analysis tools.
The ability to ask questions and receive immediate, contextually relevant answers also opens up new possibilities for broader access to data across front-office functions, reducing the need for report generation.
We've written a blog specific to this idea, which you can read here.
Generative AI also enables the creation of natural language charts and graphs that visually represent financial data using a chat-style experience. These visualisations are accompanied by explanatory text that contextualises the data.
We know that banks invest large sums of money creating management reports, regulatory reports, ESG disclosures, etc. GenAI's use to generate reports and disclosures, which are then reviewed and approved before publication, could represent the most significant transformation in report generation since the introduction of business intelligence tools.
At Revvence, GenAI will play a large part in the technology and finance transformation work we deliver for our clients. Revvence has developed a unique custom GPT application called Revvy. Based on OpenAI's ChatGPT 4o model, it is trained in finance transformation processes, best practices, and technology architectures.
Revvy is designed to understand complex topics related to finance transformation and Oracle-based technology delivery. It complements our human expertise by enabling us and our clients to review large volumes of information, including relevant documentation such as policies or regulations. It also assesses existing application codes to optimise performance and recommends improved workflows to enhance efficiency.
Revvy can be deployed to projects to accelerate every aspect of the project, including requirements gathering, application design, documentation, testing, and application code development.
Bring Your Own Model (BYOM) is a capability in Oracle’s Intelligent Performance Management (IPM) framework that allows finance teams to upload and execute custom-built machine learning models directly within the Oracle EPM Cloud environment.
Unlike built-in predictive tools that rely on Oracle’s native algorithms, BYOM allows banks to deploy models created in external environments (such as Python or R) using institution-specific logic, assumptions, and structures. These models can be invoked inside planning workflows, with predictions returned and stored in Oracle EPM for use in forecast scenarios, planning submissions, or reporting overlays.
With BYOM, banks are no longer constrained by generic models. They can build models that reflect the institution’s unique risk profile, business structure, regulatory constraints, or strategic direction. This is especially relevant in use cases like capital planning, stress testing, or internal RWA projections.
For example, a bank could bring in a custom macroeconomic forecasting model developed by its internal risk team, allowing it to generate more relevant challenger forecasts for CET1 or NII planning.
Finance teams often work closely with internal data science teams or quant modellers who create advanced forecasting logic using external tools. BYOM bridges that divide — enabling model outputs to be brought directly into the Oracle EPM platform without losing traceability, auditability, or control.
Despite being custom, BYOM models must conform to Oracle’s structured input and output format — ensuring predictions are still compatible with standard planning processes. Once uploaded, the models can be accessed and run through the IPM interface, scheduled in workflows, and reviewed alongside other planning inputs.
This allows the business to benefit from bespoke modelling without introducing shadow processes or disconnected tooling.
Predictions made using BYOM are fully auditable within Oracle EPM. Users can track which model was used, which dataset was submitted, and the resulting output. A prediction history is maintained, ensuring model performance can be reviewed over time and internal governance standards met.
Key Takeaway
Bring Your Own Model (BYOM) is a foundational capability in Oracle IPM that enables banks and insurers to integrate custom ML models into their enterprise planning processes — without compromising governance, auditability, or performance. It empowers finance teams to combine advanced, institution-specific forecasting logic with the control and structure of Oracle EPM, creating a bridge between data science and enterprise decision-making.
As financial institutions deepen their use of internal models for stress testing, regulatory forecasting, and capital planning, BYOM offers a practical path to embed that sophistication directly into the planning cycle.
The Oracle Intelligent Performance Management (IPM) suite presents a transformative opportunity for banks' finance functions, fundamentally reshaping how financial data is managed and utilised.
With its advanced capabilities—including predictive planning, automated variance analysis, AI-assisted insights, and generative AI for narrative generation and visualisation—Oracle IPM empowers finance teams to harness the full potential of their data.
By integrating advanced analytics and machine learning, the suite facilitates a more agile and proactive financial management process, enabling organisations to respond swiftly to changing market conditions and optimise their performance.
This comprehensive and innovative new capability in Oracle EPM Cloud is the north star for finance functions that want to embrace AI/ML across critical finance processes. It empowers them to create value and drive growth across the bank.
Revvence can help in several valuable ways:
Jessica P : Sep 25, 2024 4:53:08 PM
Since the launch of advanced generative AI technologies like ChatGPT in late 2022, businesses across all industries, especially banking, have found...
1 min read
PJ : Apr 14, 2025 3:48:01 PM
Banks and insurers are beginning to ask a simple but transformative question: what if you didn’t need a report to get an answer? As enterprise data...
Jessica P : Sep 12, 2024 5:39:21 PM
As global banks continue to invest in digital transformation, Oracle's Artificial Intelligence (AI) and Generative AI (GenAI) offerings, showcased at...
.png?width=1650&height=586&name=Color%20logo%20-%20no%20background%20(Smaller).png "Revvence")
.png?width=300&height=107&name=Blue%20Color%20logo%20-%20no%20background%20(Small).png "Blue Color logo - no background (Small)")
PJ