Karl Yets

In this 𝘐𝘯𝘴𝘶𝘳𝘢𝘯𝘤𝘦 𝘞𝘰𝘳𝘬 𝘪𝘯 𝘔𝘰𝘵𝘪𝘰𝘯 video, see how Indico moves submissions from intake straight into Guidewire PolicyCenter—no manual rekeying required. A high-priority submission hits the underwriting inbox. Indico ingests the broker email and attachments, structures the data, and flags a missing broker contact. Once it’s added, the file moves from “Needs Attention” to “Quote Ready” and is pushed directly into PolicyCenter—fully created and pre-filled. 𝗪𝗵𝗮𝘁 𝘁𝗵𝗶𝘀 𝘂𝗻𝗹𝗼𝗰𝗸𝘀: • Structured submissions from the start so underwriters see clean, organized data immediately. • Instant exception handling so missing information is resolved fast and work keeps moving. Seamless system handoffs so submissions flow into Guidewire Software with zero rekeying. Indico turns messy submissions into 𝗰𝗹𝗲𝗮𝗻, 𝘀𝘆𝘀𝘁𝗲𝗺-𝗿𝗲𝗮𝗱𝘆 𝘄𝗼𝗿𝗸—𝗰𝗼𝗻𝗻𝗲𝗰𝘁𝗶𝗻𝗴 𝗶𝗻𝘁𝗮𝗸𝗲 𝗱𝗶𝗿𝗲𝗰𝘁𝗹𝘆 𝘁𝗼 𝗲𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝘀𝗽𝗲𝗲𝗱𝗶𝗻𝗴 𝘂𝗻𝗱𝗲𝗿𝘄𝗿𝗶𝘁𝗶𝗻𝗴 𝘁𝗵𝗿𝗼𝘂𝗴𝗵𝗽𝘂𝘁.🚀

10

I have been working on a valuable engineering asset in my toolkit — a generic, throttling-aware Dataverse integration library built on .NET 8. Many CRM projects face recurring challenges, including: • inconsistent CRUD patterns • unpredictable throttling behavior • scattered retry logic • duplicated idempotent operations • fragmented metadata and admin utilities To address these issues, I consolidated them into a reusable, governance-grade library: Dataverse.Service. This library features: • an app-pool-based connection model that rotates across multiple application users • a unified execution policy for retries, throttling, and logging • generic CRUD, query, metadata, and bulk-operation helpers • idempotent create/upsert patterns using alternate keys • cross-cutting services for ownership, state/status, attachments, audit history, and privilege checks The goal is straightforward: to make Dataverse integrations predictable, testable, and enterprise-ready without reinventing the wheel each time. What I take pride in is not just the code but the discipline behind it: – consistent patterns – clean abstractions – DI-friendly design – host-agnostic architecture – a focus on long-term maintainability For those building on Dynamics 365 or Dataverse, moving beyond ad-hoc scripts and one-off utilities by investing in a shared foundation like this can yield significant benefits across every project. I am open to exchanging notes with anyone tackling similar challenges in CRM, Azure, or enterprise integration spaces. #AzureFunction #Azure #Dataverse #Powerplatform #Dynamics365

24

2 Comments

AI Companion vs. AI Assistant vs. AI Agent: What’s the Real Difference? Artificial intelligence is evolving rapidly, and understanding the distinctions between an AI Companion, AI Assistant, and AI Agent can help you leverage these technologies effectively. AI Companion: Your Thought Partner An AI Companion is designed to think with you — it remembers your preferences, adapts to your style, and supports your creative and strategic thinking. Remembers and adapts over time Helps brainstorm, plan, and learn Engages in natural, flowing conversations Examples: Microsoft Copilot/Chat GPT as a conversational partner, Replika-style companions for emotional support. AI Assistant: Your Reliable Helper An AI Assistant executes commands and completes tasks efficiently, responding to your instructions. Sets reminders, drafts emails, retrieves informationBoosts productivity and organization Examples: Siri, Google Assistant, Alexa. AI Agent: Your Autonomous Executor An AI Agent acts independently to achieve goals by breaking down complex tasks and taking actions on your behalf. Runs multi-step workflows Uses tools and APIs autonomously. Examples: Copilot Actions, AutoGPT-style agents. These AI types complement each other, offering powerful tools to think with, help you, and act for you.

18

1 Comment

Cyber Resilience Starts with Awareness Don’t assume your antivirus is enough. Cyber threats evolve daily. Regular training + multi-factor authentication + insurance = smarter business security. #CyberInsurance #360Cyber #Cyberresilience #InsuranceBrokers #SME #360uw

13

I advise organizations to carefully consider the following eight points before deploying Agentic AI systems. Several technical, governance and ethical safeguards are essential to ensure these systems act reliably, safely and in alignment with human goals. 1- Human Oversight Even if AI agents operate autonomously, critical decisions should still require human approval. High-risk actions (financial transactions, system changes, customer communication) should have approval workflows. Humans should be able to interrupt or override agent actions at any time. 2- Clear Permission & Access Controls Agentic AI should follow the principle of least privilege. Agents should only access the tools, APIs, and data necessary for their tasks. Role-based permissions should prevent agents from executing sensitive actions without authorization. 3- Transparent Logging & Audit Trails Organizations must maintain full logs of agent decisions and actions. This allows teams to: Track why a decision was made. Investigate failures. Ensure regulatory compliance 4- Guardrails and Policy Constraints Agents should operate within strict operational boundaries, such as: Defined task scopes: Restricted system commands. Content and compliance filters. Policy engines can ensure that agents cannot violate organizational rules. 5- Simulation & Sandboxing Before deployment, agent systems should be tested in controlled environments. Run agents in sandbox environments. Simulate edge cases and failure scenarios. Stress-test long chains of autonomous decisions 6- Continuous Monitoring Agent behavior should be actively monitored in production using anomaly detection. If an agent behaves unexpectedly: Alerts should trigger. The system should automatically pause or limit the agent. 7- Clear Accountability Framework Organizations must define who is responsible for agent outcomes. This includes: AI governance policies. Risk ownership. Incident response procedures. 8- Alignment & Safety Testing Agentic AI must undergo rigorous alignment testing to ensure it follows intended goals and does not exploit loopholes in instructions. I hope you find this helpful. #AgenticAI #ArtificialIntelligence #AIAgents #ResponsibleAI #AIGovernance #FutureOfWork #AIInnovation #TechTrends

6

Successful modernisation requires guardrails — consistent standards backed by quality gates are the foundations that lower transformation risk and keep modernisation safe while paving the way for AI. Without these standards applied, measured, and enforced across the SDLC — from requirements and analysis to architecture, design, build, test, security, release, and operations — enterprises risk fragmentation, drift, and poor outcomes. The path forward is clear: 1️⃣ Define standards - set consistent expectations across the SDLC. 2️⃣ Enable easy adoption - provide frameworks, tooling, and practices for easy uptake. 3️⃣ Measure adherence - track metrics, targets, and thresholds across the SDLC. 4️⃣ Ensure with care - apply governance functions within criticality and organisational constraints. When these foundations are in place, quality gates become more than checklists. They provide: Guardrails that ensure the right activities happen, at the right stage, at the right quality. Observability and trust — not just detecting drift, but building confidence in the process. The freedom for engineers to innovate, knowing the basics are safeguarded. And here’s where human-directed AI comes in. Quality gates grounded in consistent standards create the perfect substrate for AI augmentation. AI can: Grade outputs against standards, ensuring consistency and reliability. Alert on drift or non-adherence, helping teams course-correct early. Automate repetitive compliance checks, freeing humans for higher-value work. Leverage high quality context in context engineering to produce smarter, more relevant AI-driven decisions. That’s how modernisation, foundations, and AI intersect: quality gates as the bridge from standards to safe AI augmentation. #Modernization #SoftwareQuality #EnterpriseArchitecture #SoftwareEngineering #HumanDirectedAI #IndustrializedSDLC #QualityGates

29

4 Comments

This course is best for professionals, managers, and entrepreneurs who need to use AI (ChatGPT, Microsoft Copilot, Claude) to save time on daily operations. It is less about "how AI works" and more about "how to make AI work for you."

2

The Zambian life insurance market presents significant opportunities for digital transformation. While many processes such as client onboarding, policy administration, claims management, and customer engagement remain largely manual, technology has the potential to streamline these workflows and deliver greater efficiency, accuracy, and customer value. Automation can play a pivotal role in this transformation. Digital onboarding reduces paperwork and accelerates policy issuance. Automated reminders support timely premium payments. Streamlined claims processing minimizes delays during critical times for clients. Data-driven reporting and analytics provide actionable insights for more informed decision-making. Although I am new to the insurance sector, my background in software development and engineering has provided me with extensive experience in designing and implementing scalable systems. I see a clear opportunity to apply these principles in the Zambian context, where well-designed digital solutions can address long-standing inefficiencies and unlock growth. The future of life insurance in Zambia will depend on how effectively we leverage technology to make processes seamless, responsive, and customer-centric. This is an area where I am eager to contribute.

2

Speed up claims submission, status tracking & policy applications with AI. AI helps insurers reduce costs, combat fraud & enhance customer service through intelligent automation tailored for the region. #Insurtech #AIinInsurance #ClaimsAutomation #MENAInsurance #Arabot

11