Software Improvement Group

AI is the fastest-adopted technology in history, and it is here to stay. But, what does this mean for your organization? Are you prepared for it? What do you need to know, and where should you start? We proudly present our free and practical "AI readiness guide for organizations". Written by our very own Rob van der Veer, AI expert and author of AI standards including ISO/IEC 5338 and the EU AI Act security standard. “We want to equip organizations with the knowledge and tools they need to navigate the complexities of AI more responsibly and effectively." – Rob van der Veer, Lead Author and Senior Principal Expert at Software Improvement Group. AI readiness isn't just about technology—it’s about leadership and accountability. 💡 This guide contains 19 practical steps for leaders to implement AI in organizations, and focuses on AI governance, risk management, development, and security. Unlock AI readiness for your organization today by downloading our guide here: https://lnkd.in/e42zzX98 #AIReadiness #AIGovernance #Riskmanagement #Leadership #AI #ArtificialIntelligence

Happy Holidays from Software Improvement Group 🎄✨ May your holidays be bug-free and full of cheer! #Holidays #Christmas #NewYear

Last week, our team came together for a delicious Christmas breakfast at our office in Amsterdam. It was a morning full of holiday cheer! 🎅🎄 Our contest for the Ugliest Christmas Sweater and the Brightest Star Award was definitely one of the highlights. The creativity and spirit were off the charts—whether it was glitter, lights, or hilariously tacky holiday sweaters, everyone brought their A-game. Congratulations to the winners of this year! Ugliest Sweater Award: Mark Groot 🏆 The Brightest Star Award: Kasia Kotowska ✨ Wishing everyone a wonderful holiday season filled with joy, sparkle, and maybe even a few ugly sweaters! 🎁☃ #Christmas #Amsterdam #Holidays

With the holidays approaching it's easy to overlook that 𝐭𝐡𝐞 𝐜𝐨𝐮𝐧𝐭𝐝𝐨𝐰𝐧 𝐭𝐨 𝐀𝐈 𝐜𝐨𝐦𝐩𝐥𝐢𝐚𝐧𝐜𝐞 𝐢𝐬 𝐜𝐨𝐦𝐢𝐧𝐠 𝐭𝐨 𝐚 𝐜𝐥𝐨𝐬𝐞. If your company is operating in the EU, February 2, 2025, is a crucial date you should already be preparing for. 🇪🇺 If you haven't started yet, we're here to help. The EU AI Act is the first of its kind—a law regulating the use of AI and the first phases are coming into effect early next year. To help you navigate these significant changes and ensure you can comply with the new laws and regulations, we’re hosting a 𝐟𝐫𝐞𝐞 𝐨𝐧𝐥𝐢𝐧𝐞 𝐰𝐞𝐛𝐢𝐧𝐚𝐫. During our webinar, 𝐭𝐡𝐫𝐞𝐞 𝐭𝐨𝐩 𝐢𝐧𝐝𝐮𝐬𝐭𝐫𝐲 𝐞𝐱𝐩𝐞𝐫𝐭𝐬 will break down the most urgent regulations and a 𝐫𝐞𝐧𝐨𝐰𝐧𝐞𝐝 𝐥𝐞𝐠𝐚𝐥 𝐚𝐝𝐯𝐢𝐬𝐨𝐫 𝐰𝐢𝐥𝐥 𝐨𝐟𝐟𝐞𝐫 𝐟𝐫𝐞𝐞 𝐥𝐞𝐠𝐚𝐥 𝐚𝐝𝐯𝐢𝐜𝐞 during a live Q&A. 𝐎𝐮𝐫 𝐬𝐩𝐞𝐚𝐤𝐞𝐫𝐬: Joris Willems, Head of Technology at NautaDutilh Jaap Vink, AI Act Lead at Interprovinciaal Overleg Suzanne Galletly, Digital Skills Director at EXIN Yiannis Kanellopoulos, CEO and Founder of code4thought 𝐖𝐞𝐛𝐢𝐧𝐚𝐫 𝐝𝐞𝐭𝐚𝐢𝐥𝐬: 📅 : January 16, 2025 🕓 : 16:00–17:00 (CET) 💻 : Online (free) Don’t wait until it’s too late. Secure your spot now ⬇️ https://lnkd.in/eRUKdp2p #EUAIAct #ArtificialIntelligence #AIReadiness #AICompliance

Buy-in is a complex challenge in AI adoption, and it varies among organizations. While some express nervousness around AI adoption, others are eager for speedy integration. 🧑💻🛠️ 🤖 According to the Boston Consulting Group (BCG), a staggering 66% of leaders are dissatisfied or unsure about their organization's progress with AI. Recently, our Senior Principal Expert, Rob van der Veer, was invited to the Progression to Analog podcast, hosted by Caitlin Begg, to discuss AI readiness. When it comes to these buy-in challenges, Rob highlighted that it's important for an organization to grasp the opportunities around AI while dealing with the fears accordingly. For example, many have concerns about AI being evil or taking initiatives, but we should understand it's people with malicious intent using AI that are the issue. This is why organizations must establish clear rules and guardrails to ensure that AI's impact remains positive. To adopt a proactive approach, organizations should: 🔵 Understand the risks 🟡 Set the right policies and guidelines 🔵 Assign responsibilities By implementing the right safeguards, organizations can unlock tremendous opportunities with AI and fully harness its potential. 🔐💡 Our AI Readiness Guide is designed to help you do just that. With 19 actionable steps across four key areas, our free guide provides a roadmap to navigate AI's challenges and opportunities. Get your free guide and start preparing your organization for the future of AI today: https://lnkd.in/gK2xYDCH 🎧 You can also listen to the full podcast episode on your platform of choice: Apple: https://lnkd.in/gAFCt_gX Spotify: https://lnkd.in/gSJ-wSKk #ArtificialIntelligence #AI #AIReadiness #AISecurity

⚠️ 🌱 Pushing the boundaries of Green Software! 🌱 ⚠️ We’re proud to share that our paper “On the Effectiveness of Microservices Tactics and Patterns to Reduce Energy Consumption: An Experimental Study on Trade-Offs” has been accepted for the 22nd IEEE International Conference on Software Architecture ICSA 2025! Congratulations to Cyan Xiao, Chushu Gao (Software Improvement Group), and Justus Bogner (Vrije Universiteit Amsterdam) for this fantastic achievement! What’s it about? This study explores how applying 6 Green Architecture Patterns impacts key software quality aspects: ⚡ Energy Consumption 🚀 Performance ⚙️  Maintainability Architecture Patterns tested in the controlled experiment: - Modular Monolith vs. Microservice Architecture - Service Mesh (Istio vs. Default K8s Load Balancer) - Request Bundling - Back-end for Front-end - Caching of Read Requests - Distributing Pods over Multiple Nodes The experiment, conducted on an open-source microservice system, measured energy consumption, response time, and throughput across three load scenarios (60/120/180 TPS) using the VU Green Lab’s energy measurement infrastructure. We'll let you know when the paper is published! This study is part of the LETSGO program, a collaboration between: Schuberg Philis, Software Improvement Group, Vrije Universiteit Amsterdam (VU Amsterdam), Utrecht University, Alexandre Bonvin's Lab, Imuno Therapeutics, SURF, and Innofunding BV.

Finally! AI security in one picture. At Software Improvement Group we put together this diagram of AI security. It took a while to put this critical subject in one visual - after two years of work at the OWASP AI Exchange and for the EU AI Act security standard. In a nutshell, AI security risks consist of: ➡️ Attacks through the input of the AI model, which are unique to AI, such as evasion and prompt injection. These threats may impact the confidentiality of the model, the confidentiality of the training set and the intended model behaviour – which can be seen as the availability or integrity of a performing model. ➡️ Conventional (not unique to AI) security threats to the AI assets, (e.g. training data, input). These are threats to the system in operation or, unique to AI: to data and models in the supply chain and the development environment. Their details are covered in existing work on cyber security (e.g. forcing difficult passwords to mitigate attackers guessing passwords ). There are some attention points unique to AI though, regarding these threats and the AI assets (e.g. sensitive model input may leak into model output). These conventional threats have two types of impact: - Primary impact: the direct effect, such as stealing sensitive data by breaking into the training set database - impacting training data confidentiality. The asset is AI, but the type of impact is not unique to AI. - Secondary impact, unique to AI, which some threats have. For example: by attacking the integrity of training data, model parameters, or augmentation data (data that is added to the input), model behaviour can be changed (called data and model poisoning), or by attacking the confidentiality of the model, some attacks through the input of the model are made easier. That's it! Find all the details, including the controls to mitigate these risks, at the OWASP AI Exchange: your go-to resource on AI security, feeding straight into international standards. Help your peers by reposting or tagging, so we can attain a shared understanding on this critical topic. #ai #aisecurity

A whopping 7️⃣ 7️⃣ % of businesses are using or exploring AI. What about your IT department? 🤷♂️ Whether your organization is utilizing AI-generated code to create software systems or is developing AI models, we provide actionable insights and expert guidance to help your organization thrive. 📈 𝐀𝐈-𝐠𝐞𝐧𝐞𝐫𝐚𝐭𝐞𝐝 𝐜𝐨𝐝𝐞  AI can help programmers write code faster, but this generated code often contains bugs, security risks, and maintainability issues. ⚠️ 𝘞𝘦 𝘱𝘳𝘰𝘷𝘪𝘥𝘦 𝘦𝘴𝘴𝘦𝘯𝘵𝘪𝘢𝘭 𝘪𝘯𝘴𝘪𝘨𝘩𝘵𝘴 𝘢𝘯𝘥 𝘱𝘳𝘢𝘤𝘵𝘪𝘤𝘢𝘭 𝘵𝘰𝘰𝘭𝘴 𝘵𝘰 𝘦𝘯𝘴𝘶𝘳𝘦 𝘺𝘰𝘶𝘳 𝘈𝘐-𝘨𝘦𝘯𝘦𝘳𝘢𝘵𝘦𝘥 𝘤𝘰𝘥𝘦 𝘪𝘴 𝘩𝘪𝘨𝘩-𝘲𝘶𝘢𝘭𝘪𝘵𝘺, 𝘴𝘦𝘤𝘶𝘳𝘦, 𝘢𝘯𝘥 𝘴𝘤𝘢𝘭𝘢𝘣𝘭𝘦. 𝐀𝐈 𝐦𝐨𝐝𝐞𝐥𝐬 Our research indicates that 73% of AI systems have significant quality issues. Often, there is insufficient testing and documentation, along with prevalent security, privacy, and scalability issues. ⚠️ 𝘍𝘳𝘰𝘮 𝘦𝘯𝘩𝘢𝘯𝘤𝘪𝘯𝘨 𝘈𝘐 𝘳𝘦𝘢𝘥𝘪𝘯𝘦𝘴𝘴 𝘵𝘰 𝘈𝘐 𝘮𝘰𝘥𝘦𝘭 𝘷𝘢𝘭𝘪𝘥𝘢𝘵𝘪𝘰𝘯, 𝘸𝘦 𝘰𝘧𝘧𝘦𝘳 𝘢𝘴𝘴𝘶𝘳𝘢𝘯𝘤𝘦 𝘴𝘦𝘳𝘷𝘪𝘤𝘦𝘴 𝘵𝘩𝘳𝘰𝘶𝘨𝘩𝘰𝘶𝘵 𝘺𝘰𝘶𝘳 𝘦𝘯𝘵𝘪𝘳𝘦 𝘈𝘐 𝘴𝘺𝘴𝘵𝘦𝘮 𝘫𝘰𝘶𝘳𝘯𝘦𝘺. In the era of AI, your organization can't afford to guess. Discover our AI assurance solutions here ⬇️ https://lnkd.in/eaA6UWWH #AIAssurance #AIReadiness #ArtificialIntelligence #SoftwareDevelopment

The CSRD clock is ticking, don’t let uncertainty hold you back! ⚠️ I recently came across an article by edwin van der schoot and Marceline B. in Het Financieele Dagblad, discussing the delays in implementing the Corporate Sustainability Reporting Directive (CSRD) in the Netherlands. You can find the article here: https://lnkd.in/esTHAJkF The delay has created a "legal limbo" where companies are not formally required to comply with the CSRD. At the same time, the European Commission is considering legal action against the Netherlands and other member states that are lagging. 🤷♂️ Amid this uncertainty, some businesses are proactively following the new rules to safeguard their reputations, while others—particularly smaller companies—may choose to wait. So, what now? While these delays present challenges, I believe organizations must start viewing CSRD compliance not merely as a regulatory hurdle but as an opportunity to drive meaningful sustainable transformation. ⚙️ 🌱 For those unsure how to start, especially within IT departments, I recently wrote a blog for Software Improvement Group that outlines actionable strategies IT departments can take to align with CSRD requirements. It provides practical guidance on how Green IT can minimize the environmental impact of IT operations and support CSRD compliance. You can find it here ⬇️ https://lnkd.in/ewp-ZZDk #Sustainability #CSRD #GreenIT

New M&A Deal Update! Congratulations to DKV Mobility, a leading platform for international mobility services, on acquiring a majority stake in SMART/LAB Innovationsgesellschaft mbH, a leading e-mobility SaaS platform in Germany. Through this transaction, DKV Mobility solidifies and enhances its position as a leading e-mobility SaaS platform operator, partnering with one of Germany's most successful and prominent technology companies in the aggregation, management, and billing of e-mobility charging systems. We're happy to share that we supported SMART/LAB via IMPROVED – a global corporate finance boutique leader in the Technology, Energy, and Mobility (TEM) sectors – which acted as the exclusive financial advisor to SMART/LAB and its shareholders in this transaction, with our thorough Vendor Due Diligence. #MandA #Improved #DKV #DKVMobility #SMARTLAB #VDD #VendorDueDiligence #ITDD #DueDiligence