Bussum, Noord-Holland, Nederland
12K volgers
Meer dan 500 connecties
12K volgers
Meer dan 500 connecties
Info
As a technology leader with a career spanning over 25 years, I specialize in aligning…
Services
Activiteit
12K volgers
-
Dennis Mulder heeft dit gerepostDennis Mulder heeft dit gerepostEntering the weekend! Who you gonna call? The Waste Busters! Azure waste has a habit of roaming around like a ghost, hidden, persistent, and quietly driving up the bill. That is exactly what we help our customers tackle at the Azure Cloud Waste Company: identify waste, bust it, and keep it from coming back. And now Peter is officially part of the family helping us bust the biggest ghosts. After receiving the holy grail: the only and one Azure Waste Busting Ecto. 👻🚗. Welcome to the Waste Busters family, Peter Vader! #WasteBusters #Azure #CloudWaste #CloudEconomics #FinOps #FullCircleIT
-
Dennis Mulder heeft dit gedeeld𝐀 𝐬𝐩𝐢𝐤𝐞 𝐢𝐧 “𝐦𝐚𝐧𝐚𝐠𝐞𝐝 𝐬𝐞𝐫𝐯𝐢𝐜𝐞𝐬” 𝐜𝐨𝐬𝐭 𝐢𝐬 𝐨𝐟𝐭𝐞𝐧 𝐧𝐨𝐭 𝐚 𝐭𝐨𝐨𝐥𝐢𝐧𝐠 𝐩𝐫𝐨𝐛𝐥𝐞𝐦. It is usually an audit trail. Logging. Monitoring. SIEM. Backups. API management. When those lines suddenly grow, the invoice is often telling you one thing: Platform exceptions are multiplying, and standards are missing or optional. Because every deviation from the platform baseline quietly creates extra machinery. Here is how the “shared services” bill gets manufactured: • A workload cannot use the standard logging pattern, so you add a parallel pipeline (and pay for ingestion twice). • A team uses a different runtime or hosting pattern, so the monitoring model splits (more agents, more metrics, more dashboards, more alert noise). • One exception needs “special” audit requirements, so SIEM connectors and parsing rules multiply (and retention grows by default). • Backups are not standardized, so you end up with multiple backup products, multiple vaults, multiple restore runbooks. • API gateways get deployed per team or per domain, so you pay for duplication plus the operational overhead around it. None of that shows up as “exception cost” on the decision paper. It shows up later as: • Unallocated platform spend • Central cost centers nobody can defend • “Why is observability so expensive?” conversations • Cost optimization work that turns into endless cleanup This is why cost control is a decision-rights problem. If you want explainable spend, you need to make three things explicit: • What is the platform standard (the default, not the recommendation) • What qualifies as an exception (and who can approve it) • Who owns the ongoing run cost created by that exception Otherwise the organization does something predictable: Exceptions become architecture. And “managed services” becomes the hiding place on the invoice. Where do you see the most avoidable waste in your IT estate? How do you identify waste and eliminate it? Install the Full Circle IT DEBT Platform from the Azure Marketplace and you will know for free how much waste we find. https://lnkd.in/emMMTWwu #ITArchitecture #CloudWaste #CloudOptimization #FinOps #TechnologyLeadership
-
Dennis Mulder heeft dit gedeeldIf everything is 99.99%, nothing is actually critical. “Max availability by default” feels safe. It is also one of the fastest ways to manufacture structural cloud spend and operational complexity. Because the moment you label everything as critical, the estate behaves predictably: • Active-active becomes the reflex. • Every dependency inherits the highest tier. • Recovery becomes a promise nobody tested. • Change gets slower, because every change is now “high risk”. Availability is not a platform setting. It is a business decision with architectural consequences. Here is a practical cheat sheet I use to break “criticality inflation” in architecture discussions. Per capability, make these choices explicit. 1) Impact: what happens if it is down? • Revenue impact per hour (real number, not a feeling) • Operational impact (can people work around it?) • Reputational impact (internal, external, regulated market?) • Legal or contractual obligations (yes/no, and what exactly) 2) Time: how fast must it be back? • RTO: how long can we tolerate it being unavailable? • RPO: how much data loss is acceptable? 3) Scope: what is “the capability” really? • The user journey, not the component • The dependencies that must match the same tier • The parts that can degrade (read-only, delayed processing, manual fallback) 4) Architecture: what design matches that decision? • Single region with tested restore • Multi-zone with automated failover • Multi-region active-passive • Multi-region active-active (rarely the default) 5) Cost and operations: what are we signing up for? • Extra run cost (compute, data replication, licensing) • Extra change cost (deployments, testing, incident drills) • Extra complexity cost (data consistency, observability, failure modes) If you cannot explain the business impact, you cannot justify the tier. And if you cannot justify the tier, you are not buying resilience. You are buying comfort. How does your organization decide between cost, risk, speed, and control? How do you identify waste and eliminate it? Install the Full Circle IT DEBT Platform from the Azure Marketplace and you will know for free how much waste we find. https://lnkd.in/emMMTWwu #ITArchitecture #CloudOptimization #EnterpriseArchitecture #TechnologyLeadership #FinOps
-
Dennis Mulder heeft dit gedeeldIf your cloud bill has “data transfer” in the top 10, you are not looking at a pricing problem. You are looking at architecture evidence. Because egress and inter-region charges usually mean two things: 1) Your data does not live where you think it lives. 2) Your services are coupled across boundaries you did not design on purpose. Here is a quick cheat sheet I use to read those lines as signals. What data transfer often reveals: • A “central” data platform that every product pulls from, across regions or landing zones • Shared services in a hub that became a dependency magnet (identity, logging, API gateway, integration) • Hub-spoke that behaves like a mesh, because every spoke talks to every other spoke through the hub • Multi-region defaults that were enabled “just in case” and never revisited • Cross-zone design that quietly turned into cross-region traffic after a replatform, merger, or landing zone split Why it becomes expensive to reverse later: • Data gravity builds fast, and moving it later is disruptive • Once teams integrate across regions, the dependency graph hardens • “Temporary” routing becomes the new normal, because nothing breaks immediately Three questions to bring to the next cost discussion: • Which flows are intentional, and which ones are accidental side effects? • Which capability is allowed to be cross-region, and which one must be local by design? • Who owns the decision to keep data central, versus placing it closer to compute? FinOps can show you where the money went. But data transfer lines tell you where your architecture boundaries are leaking. What tradeoff are you currently not making explicit? Does your FinOps platform identify these problems? And do you have a process to implement the findings? Full Circle IT does just that, identify and dispose waste through tools and human intelligence. #ITArchitecture #CloudOptimization #CloudWaste #EnterpriseArchitecture #FinOps
-
Dennis Mulder heeft dit gedeeldElasticity is not free. It just moves the bill. From “cloud spend” to: Operations. Risk. And unpredictability. Many CIOs and CTOs still write “must scale” as a default requirement. As if autoscale is a neutral feature you can switch on. But elasticity is a choice with a price tag in places your invoice will not label clearly: • Reliability: more moving parts, more failure modes • Testability: you now need to test behavior across load bands, not one baseline • Incident response: capacity becomes a variable during the incident • Spend explainability: “why did it spike?” becomes a recurring meeting The hidden root cause is usually simple: No explicit capacity and performance budget per capability. So the estate does something predictable: • Everything is allowed to scale • Peak becomes the default design target • Performance becomes an unbounded expectation • The operational burden becomes “platform responsibility” A practical decision model: Elasticity earns its keep when: • demand is genuinely variable • the business value of absorbing spikes is clear • you can define safe scaling boundaries Bounded capacity is often better when: • predictability matters more than peak handling • the capability has stable demand • you need clear SLOs and a clear budget envelope Autoscale is not a requirement. It is a tradeoff. How does your organization decide between cost, risk, speed, and control? How do you discover your Azure Waste? The Full Circle IT DEBT tool identifies automatically and our Human Intelligence enforces disposition. #ITArchitecture #CloudOptimization #EnterpriseArchitecture #TechnologyLeadership #FinOps
-
Dennis Mulder heeft dit gedeeld𝐀 𝐥𝐚𝐧𝐝𝐢𝐧𝐠 𝐳𝐨𝐧𝐞 𝐰𝐢𝐭𝐡𝐨𝐮𝐭 𝐚𝐧 𝐞𝐬𝐜𝐚𝐩𝐞 𝐡𝐚𝐭𝐜𝐡 𝐛𝐞𝐜𝐨𝐦𝐞𝐬 𝐚 𝐛𝐨𝐭𝐭𝐥𝐞𝐧𝐞𝐜𝐤. 𝐀 𝐥𝐚𝐧𝐝𝐢𝐧𝐠 𝐳𝐨𝐧𝐞 𝐰𝐢𝐭𝐡 𝐮𝐧𝐥𝐢𝐦𝐢𝐭𝐞𝐝 𝐞𝐱𝐜𝐞𝐩𝐭𝐢𝐨𝐧𝐬 𝐛𝐞𝐜𝐨𝐦𝐞𝐬 𝐚 𝐰𝐚𝐬𝐭𝐞 𝐟𝐚𝐜𝐭𝐨𝐫𝐲. Most platform teams get pushed into a false choice: Standardize everything, slow teams down. Or allow autonomy, and watch the estate fragment. The waste is not in the standard. The waste is in the unmanaged exception. Because exceptions do not stay “one-time.” They create variants. Variants create permanent operational work. This is what it looks like in practice: • A separate networking pattern “just for this product” • A custom identity flow because the standard is “too slow” • A different logging setup because the default is “too expensive” • A special pipeline because the platform product is “not ready yet” Each one sounds reasonable in isolation. Together they produce predictable outcomes: • More runbooks • More on-call complexity • More security drift • More bespoke upgrades • More “nobody knows how this works” And the bill becomes harder to explain. Not because cloud is expensive. Because your operating model now funds multiple ways of doing the same thing. The fix is not “more governance.” It is a decision model that makes the tradeoff explicit: Non-negotiables (guardrails): • Security baseline • Sovereignty constraints • Identity and access model • Observability minimums • Operational ownership Controlled variation (escape hatch): • Allowed only with named owner • Explicit cost impact (build and run) • Explicit exit requirement (how do we return to standard?) • Expiry date or review moment Autonomy is not the problem. Ambiguous boundaries are. What tradeoff are you currently not making explicit? #ITArchitecture #EnterpriseArchitecture #CloudWaste #CloudOptimization #TechnologyLeadership
-
Dennis Mulder heeft dit gedeeldMost teams do not want autonomy. They want exceptions without owning the consequences. And that is how “platform guardrails” slowly turn into “platform suggestions”. The standardization vs autonomy debate is usually framed as culture. Trust. Empowerment. But in the cloud estate, it is an operating model problem. Because every exception has a run cost. And a change cost. And a security and resilience cost. Autonomy is not free. It is just rarely invoiced to the team asking for it. So the estate gets predictable behavior: • “We need a different landing zone.” • “We cannot use the standard pipeline.” • “Our workload is unique.” • “We need admin rights, just for now.” Sometimes that variation is real business value. Often it is convenience disguised as principle. A simple decision logic I use to make it explicit: Variation is allowed only if: 1) The advantage is measurable Not “faster”. Faster by how much, for which capability, and for how long? 2) The extra run and change load is accepted Who owns the on-call impact? Who funds the maintenance? Who pays when the exception blocks upgrades? 3) The exception has an expiry or a path to standard Otherwise you are not granting autonomy. You are creating permanent divergence. If any of those three are missing, it is not autonomy. It is unmanaged complexity. Guardrails are not there to slow teams down. They are there to keep the estate explainable. And to make sure variation is a deliberate business choice, not an accidental architecture outcome. How does your organization decide between cost, risk, speed, and control? #ITArchitecture #EnterpriseArchitecture #CloudOptimization #CloudWaste #TechnologyLeadership
-
Dennis Mulder heeft dit gedeeldYour Azure logging bill is often a permanent incident response setting. Not because logging is “too expensive”. Because nobody defined the default. I keep seeing the same pattern: 1 incident happens. Verbosity goes up. Retention gets extended. Extra signals get enabled “for a few days”. Useful. But the rollback never comes. So the exception becomes the new baseline. If you want to prevent Azure cloud waste, treat logging like an architectural budget per workload. Not an afterthought. A practical way to make it explicit: Define a standard logging budget per workload, with three parameters: • What do we log by default (signals and categories)? • How long do we retain it (hot vs archive)? • At what detail level do we log (normal vs debug)? Then add the part most estates miss: Explicit exception rules. • When is “incident mode” allowed? • Who can turn it on? • What is the stop moment (date, condition, or ticket closure)? • Who confirms we reverted to baseline? • Which audit requirements justify longer retention, and for which data only? This is governance. Because it forces the tradeoff into the open: cost vs visibility, speed vs control, evidence vs retention debt. Without a default, the platform will pick one for you. And “keep it all, at full detail, for longer” is a very predictable default. What tradeoff are you currently not making explicit? Call us or deploy our Full Circle IT DEBT Azure Marketplace offer to understand your logging waste. #ITArchitecture #Azure #CloudWaste #CloudOptimization #FinOps #TechnologyLeadership
-
Dennis Mulder heeft dit gedeeld𝐂𝐥𝐨𝐮𝐝 𝐰𝐚𝐬𝐭𝐞 𝐡𝐚𝐬 𝐚 𝐥𝐨𝐭 𝐢𝐧 𝐜𝐨𝐦𝐦𝐨𝐧 𝐰𝐢𝐭𝐡 𝐝𝐢𝐬𝐩𝐨𝐬𝐚𝐛𝐥𝐞 𝐩𝐚𝐜𝐤𝐚𝐠𝐢𝐧𝐠. Creating something is easy. Disposing of it properly is where the real work starts. With packaging, the hidden cost comes later: collection, sorting, recycling and disposal. And someone has to own that process. Cloud works the same way. Provisioning is frictionless: • Spin up another non-prod environment • Copy another dataset • Add another feature or experiment • Turn on additional logging “for now” • Add some extra capacity “just to be safe” A few clicks. Done. And today, finding potential waste isn’t particularly difficult either for the top services. Azure, FinOps platforms and optimization tools can all produce recommendations. But a recommendation doesn’t save a single euro. The long tail of services is a different service. We often see those services in large optimized environments still accounting for 10%+ waste. Cleaning is also a problem if you have the recommendations. Someone still has to decide: • Is this resource really no longer needed? • What depends on it? • Can we safely resize, stop or remove it? • What data must be retained, archived or deleted? • How do we prove the saving was actually realized? This is where cloud optimization initiatives often stall. Nobody wants to own the risk of deleting the wrong thing. “Temporary” becomes permanent. “Keep it just in case” becomes the safest answer. And the waste keeps accumulating. At Full Circle IT, we approach this differently. 1. We find waste continuously. Our technology and ~80 AI agents continuously analyze the Azure environment, identify anomalies, risks and optimization opportunities, and surface new insights every day. 2. Human Intelligence turns insights into decisions. Our Azure experts validate the findings in their technical and business context. We establish ownership, decision rights and deadlines, and work with the organization to actually resize, retire, archive or decommission what is no longer needed. This is the critical part: we don’t stop at recommendations. We drive disposition. 3. DEBT closes the loop. Discover. Evaluate. Balance. Transform. DEBT tracks opportunities from discovery through decision and execution to realized savings. And once the waste is removed, we put the ownership, lifecycle controls and guardrails in place to prevent the same waste from quietly building up again. Then the cycle starts again, because tomorrow there will be new waste to find. Our agent development never ends. That’s the difference between finding cloud waste and structurally eliminating it. Tools provide insight. Human Intelligence forces the decision. DEBT turns it into a continuous process. Find it. Decide it. Eliminate it. Prevent it. Repeat. That’s how Full Circle IT turns cloud waste into lasting value. DM, comment or call us if you need more information. @FullCircleIT #Azure #CloudWaste #FinOps #CloudOptimization #AzureCostOptimization #ITArchitecture
Dennis Mulder heeft gereageerd op commentaar
5 d
Exactly – autonomy without explicit ownership of the extra run, change, and resilience costs just invoices the platform team indefinitely. That's unmanaged complexity in disguise. Who in your org owns invoicing those exceptions upfront?
Dennis Mulder heeft gereageerd op commentaar
5 d
That setup screams "comfort over resilience" – until the bill arrives and ops teams drown in complexity. What's the revenue impact per hour of downtime for that capability? Without explicit business math, active/active is just expensive reflex. 😉
Dennis Mulder heeft gereageerd op commentaar
5 d
Agreed—deployment visibility without license alignment hides real compliance and cost risks. But those "unique workload" exceptions often amplify it. Tying variation to measurable value and owned ops keeps the estate explainable.
-
Dennis Mulder vond dit interessantDennis Mulder vond dit interessantSince moving back from the US to the Netherlands 4 months ago, people often ask what I miss most. My answer is always the same: the nature, the beautiful hikes. This weekend I got to relive a bit of that in 🇳🇴 Norway — an 8k hike up Pulpit Rock to recharge. It's the famous rock formation where scenes from movies like Mission Impossible were filmed, and standing at the top, it's easy to see why. Sometimes you just need to disconnect, get outside, and let the view remind you why you do what you do - and like to live on the “edge” of technology!
-
Dennis Mulder vond dit interessantDennis Mulder vond dit interessantIasa Netherlands Quarterly Event has started !
-
Dennis Mulder vond dit interessantLet's Bust some Waste!Dennis Mulder vond dit interessantEntering the weekend! Who you gonna call? The Waste Busters! Azure waste has a habit of roaming around like a ghost, hidden, persistent, and quietly driving up the bill. That is exactly what we help our customers tackle at the Azure Cloud Waste Company: identify waste, bust it, and keep it from coming back. And now Peter is officially part of the family helping us bust the biggest ghosts. After receiving the holy grail: the only and one Azure Waste Busting Ecto. 👻🚗. Welcome to the Waste Busters family, Peter Vader! #WasteBusters #Azure #CloudWaste #CloudEconomics #FinOps #FullCircleIT
-
Dennis Mulder vond dit interessantDennis Mulder vond dit interessantIk mag het eindelijk delen! Onze eerste investeringsronde is afgerond, let’s go! 🚀 Met hulp van Volta Ventures zetten we met HostYourAI de volgende stap. Onze ambitie is groot: we willen het AI-hostingplatform van Europa worden. Omdat we vinden dat we voor zoiets belangrijks als AI niet afhankelijk moeten zijn van Amerikaanse technologiebedrijven. We vinden dat mensen en organisaties zelf moeten kunnen kiezen welke modellen ze gebruiken, waar die gehost worden en wat er met onze data gebeurt. Daarvoor moeten we in Europa zelf de kennis en infrastructuur opbouwen. En dat is waar we met ons team iedere dag hard aan werken, soevereine Europese AI. Met HostYourAI bouwen we aan het AI-hostingplatform van Europa. We helpen bedrijven om open en eigen AI-modellen op Europese infrastructuur te gebruiken. In hun producten, voor hun eigen toepassingen en bij het programmeren. Wij regelen de hosting, het beheer en de optimalisatie. Met Loes bouwen we aan een Nederlandse AI-assistent. We trainen Loes modellen op Nederlandstalige data en draaien die op ons eigen hostingplatform. Wat we daarbij leren, gebruiken we om zowel Loes als HostYourAI verder te verbeteren. Met deze investering kunnen we ons team uitbreiden, het platform verder ontwikkelen en meer bedrijven in Europa helpen. Dank aan Frank Maene en het team van Volta Ventures voor het vertrouwen. En aan ons team: Daan Vielen, Hans Pinckaers, MD PhD, Margriet Jonker, Jos van Essen, Stan Jeanty, Sander Mulders, Kyra Dresen en onze klanten en iedereen die test, meedenkt en meebouwt: ik ben zo trots dat we dit samen mogelijk maken. We hebben nog zoveel plannen. Op naar de volgende stap! 💪
-
Dennis Mulder vond dit interessantDennis Mulder vond dit interessantClaude Certified Architect, Foundations. ✓ A small experiment in how far common sense and transferable knowledge can get you when preparation time is… scarce. Turns out: quite far. Know one, know most. The interfaces change. The patterns rhyme. #ClaudeCertifiedClaude Certified Architect - Foundations was issued by Anthropic to Andreea-Ioana Dan.Claude Certified Architect - Foundations was issued by Anthropic to Andreea-Ioana Dan.
-
Dennis Mulder vond dit interessantDennis Mulder vond dit interessantLast week, I was in Seattle and had the opportunity to learn more about the Microsoft Frontier Company. The overall conclusion: the next phase of AI will not be won by companies with the most pilots. It will be won by those that turn AI into real business outcomes. AI is entering its next stage: proving the value. What stands out to me is the Microsoft Frontier model: a startup mindset inside Microsoft, bringing together global sales leaders, technical leaders and deep industry expertise in one team. Because AI transformation is not just a technology challenge. It is about connecting technology to business value, putting the right governance in place, and creating an environment where teams can start, fail fast, learn and adapt. Do we have all the answers? No. But that is exactly why diversity of thought, experience and background matters. One global team, many perspectives, one mission. People who know me know that I’m energized by challenges where the path is still being shaped. And our team moving to the Microsoft Frontier Company is definitely one of them. From AI experimentation to AI impact. That is where the real work starts. 📸 AI-generated picture. Why? I was so busy enjoying the event that I barely had time to take any pictures. 😄
Licenties en certificaten
Word nu lid om alle certificeringen te bekijken
Publicaties
Projecten
-
Online Remarketing Portal
-
Global, multi-lingual portal for remarketing of DLL assets. The first such portal implemented in DLL with the purpose of improving margin and speed on sales of assets that come out of lease.
Developed in C#.NET on the Sharepoint and MS Commerce platforms.Andere bijdragersProject weergeven
Talen
-
English
Professionele werkvaardigheid
-
Dutch
Moedertaal of tweetalig
-
French
Basisvaardigheid
-
German
Basisvaardigheid
Ontvangen aanbevelingen
17 personen hebben Dennis aanbevolen
Word nu lid om dit te bekijkenOverige vergelijkbare profielen
Meer bijdragen onderzoeken
Anderen hebben Dennis Mulder genoemd in Nederland
92 anderen die Dennis Mulder in Nederland zijn genoemd op LinkedIn
Bekijk anderen die Dennis Mulder heten