AI and automation in SAM: how intelligent license managem...

It’s a Tuesday afternoon. Paul, the director of IT purchasing at a large manufacturing company, opens an e-mail that chills his blood. It’s an official letter from one of the world’s largest software vendors, politely informing him of its intention to conduct a licensing compliance audit. A quiet panic breaks out in the company. No one has a single, reliable source of truth. How many server database licenses do they actually have? And how many are in use in their distributed hybrid infrastructure, which includes three data centers and two public clouds? Are the developers in the new R&D project sure they’re using the free developer editions, or have they run expensive enterprise versions by mistake? Paul knows what awaits him: weeks, maybe even months, of painstaking manual work. Putting out fires, collecting data from dozens of incomplete spreadsheets, involving administrators who will have to manually log into hundreds of servers. He also knows how this will most likely end - a multimillion-dollar out-of-pocket expense to replenish missing licenses and potential fines.

This nightmare is a daily reality for thousands of organizations that still rely on outdated, manual Software Asset Management (SAM) methods. In today’s dynamic and complex IT environment, this approach is not only inefficient - it is a recipe for financial disaster. However, a revolution is coming. Artificial intelligence and ubiquitous automation are transforming the reactive and painful SAM process into an intelligent, proactive and automated nervous system for organizations. This article is a guide for IT, finance and procurement leaders who are tired of putting out fires. We will show how modern, AI-based SAM platforms not only protect the company from audit risk, but become a powerful tool for strategic cost optimization that, as our clients’ experiences have shown, can save up to 30% of a

ual software spending.

Why is the traditional, manual approach to SAM doomed to fail in today’s IT?

“Organizations that implement a robust SAM program can expect to reduce their software spending by up to 30% in the first year.”

— Gartner, Market Guide for Software Asset Management Tools | Source

Traditional software asset management, which was born in the era of stable, physical server rooms, was based on a few fundamental assumptions: static infrastructure, long purchasing cycles and central control. The process typically consisted of periodic, manual inventories and painstakingly comparing the results with purchasing data stored in spreadsheets. In today’s dynamic IT world, each of these pillars has fallen to ruin, rendering manual SAM completely inadequate.

1 The complexity of hybrid and multi-cloud environments: Infrastructure is no longer uniform and predictable. Applications today run in on-premise environments, in private clouds and on multiple public cloud platforms (AWS, Azure, GCP) simultaneously. Containerization (Docker, Kubernetes) and serverless architectures are adding to the mix. Manually tracking where what software is installed and running in such a distributed and ephemeral (non-persistent) environment is simply impossible.

2 The explosion of SaaS and Shadow IT software: The subscription model (Software as a Service) has democratized access to software. Any employee with a credit card can buy and start using a new application within minutes, often without the knowledge or oversight of the IT department. This phenomenon, known as “Shadow IT,” leads to uncontrolled proliferation of subscriptions, u

ecessary costs (many teams buy the same tools) and huge security and data protection risks. Spreadsheets caot keep up with these dynamics.

3. complex and constantly changing licensing metrics: Software vendors such as Oracle, Microsoft, IBM and SAP use extremely complex licensing models that are often based on metrics that are difficult to measure (e.g., number of CPU cores, PVU - Processor Value Unit). What’s more, these rules are constantly changing, and their interpretation in the context of virtualization and cloud is often ambiguous. Maintaining compliance requires specialized, constantly updated knowledge, which is lacking in most organizations.

4 Speed of change in DevOps environments: In a DevOps culture, where new environments are created and deleted automatically within minutes and changes are implemented multiple times a day, periodic, manual inventory is worthless. Data collected on Monday, is already out of date on Tuesday. SAM must operate in near real time to keep up with the pace of modern IT.

Manual SAM in today’s world is like trying to count grains of sand on a beach during a storm. It’s inaccurate, time-consuming, always late, and generates a false sense of security that ends brutally when you receive a letter a

ouncing an audit.

What is smart SAM and what role do AI and automation play in it?

Smart SAM is a fundamental paradigm shift - moving from a reactive, periodic inventory of nature to a proactive, continuous and automated software lifecycle management process. At the heart of this transformation are automation and artificial intelligence (AI), which enable tasks that are impossible at manual scale. An intelligent SAM acts as the central nervous system of an organization, collecting, processing and analyzing data in real time to provide actionable insights - specific, data-driven recommendations.

The role of AI and automation in this process can be divided into several key areas:

1 Automated Discovery & Inventory: Instead of relying on manual scans, modern SAM platforms use automated agents and API integrations to continuously and comprehensively scan the entire IT environment - from physical servers to virtual machines to containers to public cloud and SaaS applications. This creates a single, always-on source of truth about what software is installed and used throughout the organization.

2 Intelligent Data Normalization: Raw inventory data is chaotic. The same application can appear under dozens of different names (e.g., “MS Office,” “Microsoft Office 365,” “Office Pro Plus 2019”). AI algorithms, supported by huge, curated software databases, automatically clean and normalize this data. They recognize different versions, editions and packages, assigning each installation to a specific product in the catalog. This process is absolutely crucial for reliable data.

3 Automated License Reconciliation: This is the heart of smart SAM. The platform automatically compares normalized data on installed software with data on owned licenses and entitlements (which are imported from purchasing systems or contracts). The system, knowing the complex licensing rules of a given vendor, can accurately calculate the Effective License Position (ELP) - that is, answer the question, “Do we have enough licenses for the software we own?”

4 AI-Powered Proactive Optimization: This is where AI shows its greatest power. Instead of just reporting on the current state, the system uses machine learning algorithms to generate specific optimization recommendations:

“License Harvesting”: Identifies software that is installed but not used, and recommends uninstalling it and retrieving the license.
Configuration optimization: Suggests changes to the configuration of servers or virtualization clusters to minimize the need for expensive licenses (e.g., limiting an Oracle database VM to fewer cores).
Purchase recommendations: Analyzes usage trends and helps forecast future demand, suggesting the most cost-effective purchasing models.

Smart SAM is a shift from asking the question “What do we have?” to answering the question “What should we have and how can we achieve it in the most cost-effective way?”.

How does automatic discovery (discovery) lay the foundation for a credible SAM?

The “garbage in, garbage out” principle is exceptionally true in the SAM world. Even the most sophisticated AI algorithms and the best licensing analysts are worthless if they operate on incomplete or outdated data. This is why comprehensive, automated asset discovery is the absolute, non-negotiable foundation of any mature SAM practice.

The automated discovery process in modern SAM platforms is based on a multi-layered approach to provide the most complete picture of the IT environment:

1 Agent-based Discovery: A small, lightweight “agent” is installed on servers (physical and virtual) and workstations. This agent scans the machine at regular, predefined intervals, collecting detailed information about:

Hardware: Processor model, number of cores, amount of RAM (key for licensing metrics).
Installed software: List of all installed applications, their versions and editions.
Software usage: Data on which applications are actually running and how often (which is the basis for optimization). Agent-based discovery is the most precise method and provides the richest data set.

2 Agentless Discovery: In cases where agent installation is impossible or undesirable (e.g., on network devices, in secure DMZs), agentless techniques are used. A central scanner remotely connects to devices (e.g. via WMI for Windows, SSH for Linux) and retrieves the necessary data. This is less detailed, but allows coverage of parts of the infrastructure that would be invisible to agents.

3 Integrations with virtualization and cloud systems: modern SAM platforms integrate directly with the APIs of virtualization platforms (e.g. VMware vCenter, Microsoft Hyper-V) and with the APIs of public cloud providers (AWS, Azure, GCP). This allows the collection of data that is invisible from within a single virtual machine, such as:

Host-guest relationships: which VMs are running on which physical host (crucial for physical core licensing).
Dynamic movement (vMotion): Tracking how virtual machines move between hosts.
Cloud configuration information: instance types, regions, auto-scaling settings.

4 Integrations with mobile device management (MDM) and SaaS: To get the full picture, SAM platforms also integrate with MDM systems to inventory software on mobile devices and with SaaS subscription management platforms to track spending and usage of cloud applications.

With this multi-layered approach, the organization gains a single, consistent and always up-to-date repository of data about all its IT assets. This is the “single source of truth” that provides the fuel for all subsequent analytical and optimization processes. Without this foundation, any attempt to build a SAM practice is like building a house on sand.

How does AI help normalize and categorize software data?

Collecting raw inventory data is only the first step. The problem is that this data is extremely “messy,” inconsistent and difficult to interpret. The same application can be represented in hundreds of different ways, depending on how it was installed. Without an intelligent cleaning and normalization process, this data is useless for licensing purposes. This is where artificial intelligence algorithms play a crucial, though often invisible, role.

Challenge: Chaos in the raw data Imagine that automated discovery found the following entries in the company:

Microsoft Office Professional Plus 2019
MS Office 2019 Pro x64
Office 19 - professional version
Word, Excel, PowerPoint (as separate installations).
Visio 2019 Standard
Project Professional

A person is able to guess that the first three entries are the same product. He also knows that Word, Excel and PowerPoint are components of Office. But doing this manually for millions of entries is impossible. Traditional SAM systems tried to do this with static rules that quickly became obsolete.

Solution: normalization based on AI and huge knowledge bases Modern SAM platforms, such as Flexera One, use a combination of AI and gigantic software knowledge bases curated over years. The normalization process works as follows:

Comparison with a directory: Each raw entry is compared with a catalog containing millions of software signatures. This catalog contains information on all known manufacturers, products, versions, editions and packages.
Pattern recognition and heuristic rules: AI algorithms, trained on millions of examples, can recognize patterns in names and automatically match even unusual entries to the correct product. They can understand, for example, that “MS” is short for “Microsoft” and that “x64” means the 64-bit version.
Package identification (Bundling): The system automatically recognizes that individual installations of Word, Excel and PowerPoint are part of Microsoft Office. Instead of counting them as three separate products, it bundles them into one item, which is crucial for correct license billing.
Data categorization and enrichment: Once normalized, data is enriched with additional metadata from the catalog, such as:
Software category (e.g., database, operating system, office application).
End-of-Life date of support, which is important for risk management.
Information about security vulnerabilities associated with a particular version.
License tags available.

Through this process, a chaotic list of raw data is transformed into a clean, structured and reliable list of software assets. Instead of thousands of incomprehensible entries, the SAM manager sees, for example: “Microsoft Office 2019 Professional Plus - 1250 installations. ” Only on such prepared data can license reconciliation and optimization processes be run. The automation of this step is one of the biggest breakthroughs that AI has brought to the SAM world.

How does machine learning optimize license usage (license harvesting)?

One of the biggest sources of waste in IT is software that the company pays for, but that no one uses. These are licenses installed on the computers of employees who have changed departments, left the company, or simply stopped using a particular application. The process of identifying and recovering these licenses, called “license harvesting,” is one of the fastest ways to generate real savings. Traditionally, it was a manual and inefficient process. Today, thanks to machine learning, it is becoming automated and extremely precise.

How it works. Modern SAM platforms, equipped with monitoring agents, collect information not only about what is installed, but also how it is used. The agent is able to track whether a particular application process is running, how long it has been active and whether the user is actually interacting with it. These huge sets of usage data are then analyzed by machine learning (ML) algorithms.

Defining usage thresholds: the SAM administrator defines policies that define what constitutes “unused” software. For example: “Software is considered unused if it has not been run even once in the last 90 days” or “if the total time of its active use in the last quarter was less than 2 hours.”
Pattern analysis by ML: The ML algorithm analyzes historical data for each user and each application. It can distinguish regular, daily use from occasional, one-time use. It identifies software that meets the criteria of “unused.”
Generating recommendations: The system automatically generates a list of candidates for “license recovery.” The report may look as follows:
“Visio Professional application is installed in 300 users, but 85 of them have not run it for more than 6 months. Potential savings: X PLN.”
“The license for the Adobe Creative Cloud suite is assigned to 50 users from the marketing department, but usage data shows that 10 of them use Photoshop exclusively. Recommendation: change their plan to a cheaper one that includes only one application. Potential savings: y£.”
Automated recovery process: In mature implementations, this process can be partially or fully automated. The system can:
Send an automatic notification to the user: “We’ve noticed that you’re not using the X app. Do you still need it? If you don’t respond within 14 days, it will be automatically uninstalled.”
Create a ticket in an IT Service Management system (such as ServiceNow) with the task of uninstalling the software.
In case of integration with software distribution systems (e.g., SCCM), automatically start the uninstallation process.

Benefits:

Direct financial savings: Recovering each unused license represents real, measurable savings that can be allocated to other investments.
Security risk reduction: Every unused and un-updated application is a potential gateway for cyber attacks. Removing u

ecessary software reduces the attack surface.

Optimize future purchases: Actual usage data is an invaluable source of information when renegotiating contracts with suppliers. Instead of buying licenses “for stock,” the company can make decisions based on real demand.

License harvesting based on ML is an excellent example of how smart SAM is moving from passive reporting to active value generation for the business.

What is SaaS spend management (SaaS) and how does AI help combat shadow IT?

The explosion of Software as a Service (SaaS) applications has created an entirely new and extremely difficult to manage category of spending and risk. Unlike traditional software, SaaS applications are not “installed” - they are subscribed to, often by employees themselves using a credit card, bypassing the IT department. This leads to chaos, known as “SaaS sprawl” and “Shadow IT.” SaaS Spend Management is a key new discipline within SAM, with AI playing a revolutionary role.

SaaS challenges:

Lack of central visibility: IT and finance often have no idea how many SaaS applications are in use at a company. Studies show that the average large company uses several hundred different SaaS applications, of which IT knows about less than half.
Waste and duplicate subscriptions: Different teams, unbeknownst to each other, buy the same or very similar tools (e.g., three different project management applications). Employees leave the company, but their paid subscriptions are not canceled.
Security and RODO compliance risks: When employees process company and customer data in unauthorized SaaS applications that have not been verified by the security department, it creates a huge risk of data leakage and breach.

How does AI help you regain control? Modern SaaS management platforms, often part of integrated SAM platforms, use AI to automatically discover and rationalize this chaos.

Automated discovery through analysis of financial data: Instead of asking people what applications they use (which is inefficient), these systems integrate directly with a company’s financial and accounting systems. AI algorithms analyze expense data, credit card payments and invoices, automatically identifying all recurring payments to SaaS providers. They can recognize that “Atlassian Pty Ltd” is a payment for Jira, and “Slack Technologies” is a payment for a messenger.
Identification of owners and usage: the system correlates financial data with data from HR systems to identify which employee or department owns which subscription. What’s more, by integrating with Single Sign-On (SSO) systems (e.g., Okta, Azure AD), the platform can collect actual login and usage data for individual applications.
Intelligent rationalization and optimization: Based on the collected data, AI generates specific recommendations:
“We have identified 15 active licenses for Tool X assigned to employees who have left the company. Recommendation: cancel immediately. Savings: X PLN per year.”
“The company uses three different video conferencing tools: Zoom, Teams and Webex. Recommendation: standardize on one tool and renegotiate the enterprise contract. Potential savings: y£ per year.”
“User John Smith is assigned a premium license for tool Y, but usage data shows that he uses only basic functions. Recommendation: downgrade to a standard license. Savings: z£ per month.”

Thanks to AI, SaaS management is no longer a battle with windmills. It’s becoming an automated, data-driven process that allows companies to regain control, reduce waste and minimize the risks associated with uncontrolled cloud application sprawl.

How to build a business case (business case) for investing in a smart SAM platform?

Investing in a modern, AI-based software asset management platform is a significant expense. In order to get the board’s support and budget, it is necessary to prepare a solid business case (business case) that clearly demonstrates that this is not just an “IT cost,” but a strategic investment with a high return (ROI). This justification should be based on three pillars: cost reduction, risk minimization and increased operational efficiency.

Pillar 1: Cost Reduction (Hard Savings) This is the most appealing argument. You need to estimate the potential measurable savings that the implementation of the platform will bring.

Savings from license optimization: As industry analysis (e.g., Gartner, IDC) and ARDURA customer experience show, companies that implement mature SAM practices can reduce their a

ual software spending by 15% to 30%. Take your current a

ual software budget (on-premise and SaaS) and calculate how much the potential savings will be (e.g., 20% of £10 million is £2 million per year).

Eliminate waste in SaaS: As mentioned earlier, up to 30% of SaaS spending is wasted. Assess this value for your company.
Avoid overpaying when renegotiating: Data from the SAM platform gives a huge negotiating advantage when renewing contracts with suppliers. Instead of relying on supplier data, the company has its own reliable information about actual demand.

**Pillar 2: Risk Mitigatio ** This pillar is more difficult to quantify, but equally important. It is about avoiding potential, often huge, costs.

Avoiding audit penalties: The cost of a licensing non-compliance detected during an audit can run into the millions. The SAM platform is the best insurance policy against this risk. Provide market examples of companies that have paid large fines.
Security risk reduction: Identifying and removing unauthorized or outdated software reduces the attack surface and the risk of security incidents, the cost of which (data loss, downtime, loss of reputation) can be catastrophic.
Ensuring regulatory compliance (e.g., RODO): Software controls, especially SaaS, are key to ensuring that personal data is processed in a lawful ma

er.

Pillar 3: Increasing Operational Efficiency (Soft Savings) This is about saving time and human resources that can be allocated to more valuable tasks.

Automate manual processes: Estimate how many work hours per year your employees (admin, finance, purchasing) spend on manual data collection, report creation and audit preparation. The SAM platform automates 90% of these tasks.
Faster decision-making: Access to reliable, real-time data allows for much faster and better purchasing and strategic decisions.
Democratization of data: Making cost and usage data available to managers and teams allows for better decision-making at all levels of the organization.

ROI calculation: Finally, weigh all these benefits against the cost of investment (cost of SAM platform license, cost of implementation, possible consulting support). The formula for ROI is: (Total Benefits - Total Cost) / Total Cost. For mature SAM implementations, ROI is typically achieved within 6-12 months, making it one of the most cost-effective IT investments.

What are the stages of implementing a mature SAM practice in an organization?

Implementing a mature, AI-based SAM practice is not a one-time software installation project. It’s a transformational program that involves technology, processes and people. As with other mature practices, its implementation should proceed in an evolutionary, step-by-step fashion, building the foundation and gradually expanding in scope and value.

Stage 1: Foundations and Visibility (Months 1-3) The goal of this stage is to build a single, credible source of truth about IT assets.

Establish a team and gain support: Establish an interdisciplinary project team (IT, finance, procurement) and secure support from a board sponsor.
Platform selection and implementation: Selection of a suitable SAM platform (e.g., Flexera One) and implementation of its technical components (e.g., installation of discovery agents).
First inventory: Run automated discovery processes and collect raw data on all hardware and software assets.
Loading of entitlement data: Importing data on owned licenses from contracts and purchasing systems.
Achieving visibility: At the end of this stage, the organization should have, for the first time, a complete, consolidated picture of what it has and what it has installed.

Stage 2: Compliance and quick wins (Months 4-9) In this stage, we focus on achieving compliance with key suppliers and generating the first quick savings.

License Reconciliation: Run processes to automatically reconcile installation data with license data to calculate Effective License Position (ELP) for the top 2-3 most important and highest risk vendors (e.g., Microsoft, Oracle).
Repairing non-compliance: Take action to fix identified licensing gaps (e.g., uninstall software, purchase missing licenses).
“License Harvesting: Launch the first recovery campaigns for unused software to generate quick, measurable savings.
Achieving compliance: The goal is to achieve a state where the company is ready for an audit from key suppliers at any time.

Stage 3: Optimization and Integration (Months 10-18) Now that the foundation is solid, you can move on to more advanced optimization activities and integration of SAM with other IT processes.

Server license optimization: Use advanced platform features to optimize licenses in virtual and cloud environments (e.g., by optimizing cluster configurations).
SaaS Spend Management: Expand SAM to include SaaS subscription management, Shadow IT identification and application portfolio rationalization.
Integration with ITSM and FinOps: Integ rate the SAM platform with systems such as ServiceNow (to automate license recovery processes) and FinOps practices (to get a complete picture of costs in the cloud).

Stage 4: SAM as business intelligence (continuous process) At the highest level of maturity, SAM ceases to be just a control function and becomes a strategic data source for the entire organization.

Forecasting and scenario planning: Use historical data to accurately forecast future software demand and model the costs of various scenarios (e.g., migration to the cloud).
Support for IT strategy: Data from SAM is used to make strategic decisions, such as standardizing technology, planning the application lifecycle and negotiating global vendor agreements.
Continuous improvement: The SAM process is regularly measured, analyzed and improved.

Such a phased journey allows for a gradual building of competence, proving value every step of the way and ensuring that SAM implementation will be a lasting success and not just a short-term project.

What does the software asset management (SAM) maturity model look like?

Organizations, like people, go through different stages of development in their journey toward excellence in software asset management. Understanding which stage your company is at is key to planning a realistic and effective strategy for the future. The table below shows the SAM maturity model, which can serve as a diagnostic tool and roadmap.

Maturity stage	Characteristics	Key processes	Technology	Main objectives
Stage 1: Reactive (Ad-Hoc).	No central control. Knowledge of licensing is fragmented and tribal. Action is taken only in response to an audit.	Manual collection of data on demand. Firefighting.	Spreadsheets, emails, knowledge of individual administrators.	Survive the audit. Minimize financial penalties.
Stage 2: Managed (Centralized Inventory).	There is a central repository of asset data. An inventory tool has been implemented. The IT team has basic visibility.	Periodic, automated scanning. Basic reporting on installed software.	Dedicated discovery and inventory tool. Central database (CMDB).	Building a single source of truth. Answering the question, "What do we have installed?"
Stage 3: Proactive (License Optimization).	The organization regularly reconciles holdings with licenses. Optimization processes have been implemented.	Continuous license reconciliation (reconciliation). "License harvesting". Basic SaaS management.	Integrated SAM platform with software catalog and licensing engine.	Achieve and maintain compliance. Generate savings through optimization.
Stage 4: Strategic (SAM as business intelligence).	Data from SAM is a key source of information for strategic decisions in IT and business. SAM is integrated with FinOps and security.	Forecasting, scenario planning. Software lifecycle management. Continuous improvement.	SAM platform integrated with financial, HR, ITSM and security systems. Use of AI and predictive analytics.	Maximize value from software investments. Supporting digital transformation strategies.

Planning an IT project? Learn about our Software Development services.

Let’s discuss your project

Have questions or need support? Contact us – our experts are happy to help.

How does ARDURA Consulting’s expertise maximize the return on investment in SAM?

Implementing a modern SAM platform is a powerful step, but technology alone is not enough. Maximizing the return on this investment requires deep, specialized knowledge of complex licensing rules, experience in leading transformation programs, and the ability to connect technical data to business objectives. This is where ARDURA Consulting delivers unique value.

As a Trusted Advisor, we are not just a software reseller. We are your strategic partner at every stage of your journey to SAM maturity. Our expertise allows you to accelerate and multiply the benefits of your investment.

Strategy and Implementation: We help you design and implement a SAM program from the ground up - from building the business case, to selecting and implementing the right platform, such as Flexera One, to defining and implementing new processes in your organization.
In-depth licensing knowledge: our team of experts has years of experience interpreting the complex licensing agreements and metrics of major vendors such as Microsoft, Oracle, IBM and SAP. This knowledge helps avoid costly mistakes and maximize savings.
Support in audits and negotiations: We act as your support during license audits, helping you prepare data and conduct discussions with the supplier. Our experience and reliable data from the SAM platform strengthen your negotiating position during contract renewals.
Integration with FinOps and cloud services: We understand that SAM is a key component of a broader IT financial management strategy. Our expertise in FinOps and cloud services allows us to create a cohesive, holistic system for managing all technology spending.
Flexible support: As part of our flexible collaboration models, we can provide dedicated SAM analysts or managers to support your team operationally, accelerating the achievement of your goals.

Investing in a smart SAM is one of the most cost-effective decisions an organization can make. Working with ARDURA Consulting is a guarantee that this investment will be made in an efficient ma

er, and that the results - in the form of real savings and minimized risk - will come sooner than you expect.

If you are ready to turn licensing chaos into a source of strategic advantage, consult your project with us. We’ll help you unlock the full potential of intelligent software asset management.

AI and automation in SAM: how intelligent license management can reduce your IT costs by 30%?