Bryan Guffey

Leading change across an MSP is equal parts project management, patience, and problem-solving. Right now I’m helping drive a major shift: migrating our client base from Syncro to ConnectWise RMM. On paper, it’s “just” an RMM move. In reality, it’s touching everything: agent deployment strategy, policy baselines, monitoring standards, patching expectations, alert noise reduction, documentation, and (most importantly) maintaining trust while change is happening behind the scenes. At the same time, I’m working across a wide mix of firewall brands and environments. Each has its own way of handling policies, objects, NAT rules, and the little “gotchas.” What’s been most interesting is how much the success of an RMM migration depends on getting the fundamentals right at the edge: consistent security policies, clear exceptions, and predictable traffic behavior so the tools can actually do their job. A few reflections I’m taking from this: • Standardization is powerful, but every client still has “one exception” that tests your process. • The best migrations aren’t rushed. They’re repeatable, documented, and easy to validate. • Communication matters as much as configuration. A calm, confident rollout beats a perfect plan no one understands. Big changes like this are a reminder of why I enjoy MSP work: no two environments are the same, and the work you do today can make every support ticket tomorrow easier. If you’ve led (or survived) an RMM migration, I’d love to hear what made the biggest difference for you: process, tooling, communication, or something else? As well as any tips from you ConnectWise gurus! I would love any insight you may have to give. #MSP #ITSupport #RMM #ConnectWise #Syncro #CyberSecurity #Firewalls #NetworkEngineering #ITOperations

3

1 Comment

Show me your error logs and I'll tell you who really runs your company. Not the org chart. Not the leadership team. The actual power structure. Thirty years of building software and running servers: errors aren't random. They cluster around dependencies. A 500 error means someone needs someone else's data. A timeout means a handshake nobody documented. Database locks reveal competing priorities fighting for the same resource. Which errors get fixed immediately and which live for months? That's your real org chart. The auth service goes down and everyone scrambles. The reporting API times out and nobody notices for weeks. Because executives need dashboards. Nobody reads the reports. We fix what powerful people depend on. Everything else is "on the backlog." Your error logs don't lie about priority. Your roadmap does. Want to know who has real influence in your company? Find the service that, when it goes down, makes everyone else stop working. That's your single point of failure. Both technically and politically. #DevOps #OrganizationalCulture #TechnicalDebt #Leadership #SystemsThinking

12

1 Comment

Security Architect Q&A: Answering Your Security Architect Career Questions Security architect careers—LIVE Q&A with Mike Gibbs (CEO, Go Cloud Careers). In this free session, Mike breaks down what a security architect does, how to become a security architect, and the security architect skills you need to build a standout cloud security architect career and cybersecurity architect career. #SecurityArchitect #SecurityArchitectCareer #CloudSecurityArchitect

5

Partnerships, acquisitions, architecting truly integrated solutions to solve defined, tactical challenges instead of building product marketing slides. It takes a village of solid customer engagement strategies, actionable, accessible clean data, intelligence, reliable, compliant, scalable, redundant infrastructure and trusted human relationships.

4

Every VP role has infrastructure they need to own. But if you don't define it upfront, they'll spend 6 months asking permission instead of building. I see this too often. Strong VP. Great background. Right stage fit. But nobody defined what they actually own. So they wait. They ask. They schedule meetings for clarity. Six months in, nothing has moved. You can hire a rockstar VP, but if you don't clarify what they own, they'll never perform like one. Answer these questions: Which systems? CRM, billing, data warehouse, CS tools. Which processes? Quote-to-cash, onboarding, renewals, forecasting. Make it clear what they own. Then let them run it. Read the full framework: https://lnkd.in/gVctPw77 Next: Metrics and decision rights.

17

6 Comments

Outcome-Focused Ops: How to Lead When Things Go Wrong When an upgrade hiccups in prod, the win isn’t finding someone to blame; it’s aligning the room on the end state and moving there step by step. I share a simple playbook to use in real incidents and post-upgrade surprises. Highlights: • Set a clear end state: restore service, protect data, reduce repeat risk. • Run with accountability, not blame: small, reversible steps with named owners. • Communicate with cadence: timestamped “Now / Next / Risks” updates. • Finish strong: convert fixes into automation, tests, and tighter runbooks. If you lead ops, SRE, or ERP teams, this is a practical checklist you can use tonight or at 2 a.m. next week. Read it here: https://lnkd.in/g_8sjhvC #Leadership #ITOps #SRE #IncidentManagement #ChangeManagement #DevOps

4

2 Comments

Tuesday Thought: Why Most Proposals Struggle Before They Even Start Here’s something I’ve seen over and over again across federal pursuits… Teams wait too long to bring architecture into the conversation. By the time solutioning starts, the RFP is already being “answered” instead of being shaped. And that’s where things quietly go wrong. Because the best proposals don’t just respond to requirements they frame the narrative early. They: Define the art of the possible Align technical vision to mission outcomes Reduce downstream friction in writing, pricing, and staffing And most importantly… make the story almost write itself When architecture leads early: 👉 The compliance matrix becomes intentional 👉 The BOE becomes defendable 👉 The win themes become obvious When it doesn’t: You get disjointed sections, forced messaging, and a scramble during Red Team. I’ve found that the difference between “good” and “winning” proposals isn’t effort… It’s when and how the solution is shaped. Curious how others see it. Where do you bring architecture into your capture/proposal lifecycle? Early shaping… or after the RFP drops? #SolutionsArchitecture #ProposalStrategy #CaptureManagement #FederalIT #WinningWork

1

Steps to Configure Custom License Reclamation Policies. They need access governance that lives where requests, approvals, and changes already happen. If you’re running Jira Service Management with Okta, Entra, or Google, you can design real least privilege without slowing anyone down.

1

I don't think we celebrate the engineers at managed and professional services companies enough. Ya'll are rock stars. An engineer at a single company might become a world-class expert in 𝘵𝘩𝘢𝘵 environment. They know every quirk, every legacy decision, every weird workaround from 2019 that nobody documented. But a services engineer? They walk into chaos they've never seen before - every single week. Different architectures. Different tech stacks. Different levels of technical debt. Different "we tried to fix it ourselves first" situations. They don't get months to ramp up. They get hours. Sometimes minutes. And they have to diagnose problems without the institutional knowledge that internal teams take for granted. No one's going to say "oh yeah, that always happens when Jenkins does the thing." They have to figure out that Jenkins does the thing, that it's always done the thing, and that the thing is actually three things. That pattern recognition? You get it from being thrown into the deep end of a hundred different pools and learning to swim in all of them. So yeah. Services engineers. Wildly under-celebrated. A big thank you to all of our engineers at Rhythmic Technologies for doing what you do.

29

2 Comments

Want a CISSP-certified workforce without slowing down operations? Learn how to upskill your cybersecurity talent without disrupting workflows. From eligibility to exam prep to enterprise best practices, this guide covers it all. Build resilience, credibility, and stronger security leadership. Read  the blog at https://lnkd.in/gC6tfcsh #Cybersecurity #CISSP #EnterpriseSecurity

7

A lot of you asked what this pod actually does. Here's the honest answer for this week. I finished building the orchestrator — the Chief of Staff agent that plans, delegates, and gates everything before it reaches me. The first job I'm pointing it at isn't the weekly status rollup I sketched in earlier posts. It's something smaller, more honest, and more universal. Meeting prep. The pod's first real test is the meetings I already have on my calendar this week. I prompt it with the meeting name, date, and attendees. Chief of Staff restates the plan back to me before delegating — that intermediate step has already caught two scope mistakes I would have let slide. Analyst pulls open items, recent activity, and any incidents relevant to the people in the room. Scribe drafts a 1-page brief in my voice. Chief of Staff presents it back to me — approve, edit, or cancel. Why meeting prep first. I almost led with the weekly rollup because it's the more impressive artifact. I changed my mind for two reasons. One: it's harder to be honest about a weekly cadence when the agent has only existed for a few days. "Did it work this week?" becomes a question I can't yet answer five times in a row. Two: meeting prep is closer to how most people actually use their morning. Everyone preps for something. If the seams hold here — if the Chief of Staff plans cleanly, the Analyst returns clean facts, the Scribe drafts in my voice, and the approval gate actually saves me from sending something half-baked — that's the proving ground. The rollup is the codification, later. It earns its place by inheriting patterns this workflow has already proven. What I'm watching for over the next two weeks. — Does the Chief of Staff's plan-before-action step keep catching real mistakes, or does it become noise? — Do Analyst's findings stay structured and cited, or drift toward summary? — Does Scribe's voice hold across meetings with very different audiences (a 1:1 vs. a vendor pitch vs. a CIO update)? — And the one I care about most: do I actually approve fewer of these unedited as the meetings get higher stakes? Five consecutive clean runs is the bar I set. I'll tell you what happens when I hit it — or when I don't. #AIAgents #AIArchitecture #BuildingInPublic

16

🚨 Attention All Contract Software Engineers & IT Consultants 🚨 As someone who has been in the IT Staffing & Services industry for over 25 years, I feel it's important to raise awareness about a growing concern in our industry. 🔻 We’re seeing a trend where some vendors (especially preferred vendors) are reducing consultant bill rates and extending payment terms (Net 60, Net 90) — claiming it's due to "client-side changes." But here’s the truth: ❗ In most cases, these reductions are NOT coming from end clients. They are being implemented unilaterally by intermediary vendors, often under pressure or for their own margin protection. This is not a market correction, it’s manipulation driven by desperation. ⚠️ Consultants, please be cautious: ✅ Always verify any rate change directly with your end client or implementation partner. ✅ Clarify payment terms before signing any MSA or onboarding document. ✅ Be wary of sudden Net 90 or "pay when paid" terms — they can trap you. ✅ Keep proper documentation and communication trail. ✅ When in doubt, ask questions — your livelihood depends on it. This is a critical time in the market. Don’t let panic or miscommunication cost you your income or put you at legal risk. Protect your value, your time, and your trust. If you need any advice or guidance, feel free to reach out. #ITStaffing #ConsultingLife #RateReduction #Net90 #VendorScams #Awareness #SoftwareEngineers #ITConsultants #W2 #C2C #TechJobs #ContractorsRights

8

Workday: "Why Didn't I Think of That?": Static Email Notification Sometimes there are simple solutions to frequently recurring challenges. Sometimes you still need a Neo in the Workday Matrix. For those simple solutions that I see a need for more often, I'll be posting occasionally under "Why Didn't I Think of That?" Today's post is on Static Email Notifications, specifically for Distribution Lists. Ever felt the joy of configuring Workday notifications, only to hit a snag? 😩 You nail the security groups, but then comes the curveball: a trusty old **static email Distribution List** that doesn't quite fit the security mold. Creating a whole new security group just for that? Seems a bit overkill, right? So, you go for the simple route: sending a Business Process (BP) notification to that static email. Easy peasy... until you try to include some *configured content* in that email. Cue the Workday notification equivalent of the Blue Screen of Death: the dreaded **"[not available]" tag**! 😱 Imagine this hitting your Accounts Payable team: "Hello Accounts Payable team! **[not available]** has been terminated effective **[not available]**. The employee ID is **[not available]** so please be sure to cancel all transactions related to **[not available]** before his **[not available]** is **[not available]**. Remember, we’re all **[not available]** for you." Turns out, Workday keeps its notification content under lock and key, secured to the recipient. When it sees a static email, it's like, "Who dis?" and throws up its hands (or rather, its "[not available]" tag). And the plot thickens! What if you want that termination notification to land on the AP team's virtual doorstep *on* the effective date, not a moment sooner? Well, with static emails, you're fresh out of luck on the **effective-dating front**, too. So, you can't include secured content, AND you can't effective date those notifications when using static email addresses... or can you? Check out the 7 simple steps below to solve it!  😉 #Workday #HRTech #Notifications #TechTips #Engagement

76

8 Comments

Every IT Project Manager has heard that "can we just..." line. Usually, "one small change" is anything but small. It means another round of testing, updating the documentation, and dealing with whatever breaks downstream. I have stopped just saying no, though. Instead, I lay out the impact. Once stakeholders see the trade-offs, such as the impact on the timeline, budget, or risk level, they usually make a smarter call. That's not being difficult; it's just responsible delivery. How do you explain scope impact to non-technical stakeholders?

1

User asks: "How many PTO days do new employees get?" Your single-store RAG retriever returns 4 chunks: one HR policy, one refund clause, one termination paragraph, one webhook config. The LLM cheerfully blends them. The answer sounds great. It's also wrong. That's context pollution — and it's what kills most RAG solutions. The fix is simple! Separate your knowledge into multiple stores, and route to the right one. The solution looks like this: → One VectorStore per domain (FAQ, Legal, Tech, HR) → One shared embedding model (don't get clever here) → One cached QuestionAnswerAdvisor per store, built in @PostConstruct → A router decides which advisor to use — then it's the same .advisors(...).user(...).call() recipe Just published Post 8 of my Spring AI RAG series — Multi-Document RAG: Multiple Collections, Smart Routing, and Cleaner Top-K. Read the full post: https://lnkd.in/duvRz_Ds #SpringAI #RAG #Java #SpringBoot #LLM #AIEngineering #VectorDatabase

1

When I started building the talent pipeline for Virtrify - Build Virtual Teams, I made a decision that most people in this space wouldn't think to make. I went looking for engineers...not software developers. Real engineers. People trained to think in systems, optimize processes, and break down complexity into components that can be rebuilt better. Venezuela has produced them at scale. The engineering education system there is among the most rigorous in Latin America. These are people who were trained to solve hard problems with limited resources...which, if you think about it, is exactly what AI implementation demands. You're not always working with perfect tools or clean data. You're figuring it out inside a real business with real constraints. What I found is that the AI layer...the tools, the platforms, the workflows...can be taught. What cannot be taught is the way a trained engineer approaches a broken system. The instinct to diagnose before building. The patience to understand before automating. The discipline to document so someone else can maintain it. That mindset transfers directly into this role. Every time. Venezuela didn't just give us a talent pool. It gave us exactly the profile we were looking for...and we weren't even sure it existed until we found it. The insight: Venezuela has one of the highest concentrations of formally trained engineers per capita in the region. That's not an accident. That's decades of institutional investment in technical education. We're just the ones who figured out how to connect it to the fastest-growing need in the global market. #Venezuela #HireVenezuelan

11

Hiring isn’t just about filling roles — it’s about making the right structural bet. FTEs vs Contractors becomes even more critical when dealing with niche skills like ServiceNow. Here’s a simple lens 👇 🔵 Hire FTEs when: - You’re building core capability & platform ownership - The skill is strategic and recurring - You need long-term accountability + knowledge retention 🟠 Use Contractors when: - You need speed and flexibility - The skill is specialized or short-term - You’re scaling for projects or transformation spikes ⚖️ The reality? It’s not either/or. The best organizations: - Keep architecture & governance in-house - Use contractors to accelerate execution and bring niche expertise 💡 Rule of thumb: Strategic + recurring → FTE Specialized + episodic → Contractor 🚨 Biggest mistake: Outsourcing your brain and keeping only execution. Because in platforms like ServiceNow, who owns the platform… owns the future. #ServiceNow #HiringStrategy #Leadership #WorkforceStrategy #DigitalTransformation

49

4 Comments

I’ve noticed a lot of teams are stuck in a "Lethal Trifecta" right now. They’re running unmanaged AI agents with broad permissions in local environments. It’s basically "vibe coding" your way toward a security disaster. Tonight in San Francisco, we’re moving past the policy talk to look at the actual infrastructure needed to ship agents safely. If you’ve been trying to figure out how to give agents the context they need without giving away the keys to the kingdom, come through. Details in the comments.

26

2 Comments