Tech for Business

On the Tech for Business podcast, Collin (Manager of Solutions Engineering at ThreatLocker) and Nate (Director of Cybersecurity at CIT) discuss how to build cybersecurity awareness and secure habits earlier—especially in K–12 and universities—based on Colin’s dark web presentations showing frequent compromises in education. They explore generational shifts as kids adopt internet-connected devices earlier, the need to normalize fundamentals like password management, passphrases, password managers, and MFA/2FA (even for gaming accounts), and the growing importance of social engineering and spotting AI deepfakes. The group debates who should own curriculum standards, notes North Dakota’s requirement for at least one computer science or cybersecurity course to graduate, and considers “digital citizenship,” parent responsibility, and the role of schools, industry, and regulation. They also discuss improving workplace onboarding and training through better access controls and aptitude-based, personalized education.

00:00 Meet the Guests

00:46 Why Youth Cyber Matters

03:53 Campus Reality Check

05:56 Kids Tech Skills Gap

08:33 Gaming as Security Hook

10:51 Curriculum Needs Updating

13:26 Digital Citizenship Debate

15:12 Parents Schools and Equity

17:34 Regulation and Age Checks

21:45 Core Fundamentals to Teach

22:18 Passphrases Basics

22:40 Deepfakes And Social Engineering

24:26 Social Media Risks For Kids

25:53 Spotting Tech Skill Gaps

27:25 Curiosity As A Tech Signal

28:42 Parenting Misinformation Lessons

30:23 Rethinking Security Training

33:17 Tools Over Awareness Alone

36:15 Aptitude Tests And Personalization

40:24 Curriculum And Who Decides

41:08 Big Tech Shaping Education

43:51 Wrap Up And Next Episode

Kelsey (AI Operations Coordinator at CIT) and Kyle (President & CEO) explain how they built an AI “morning briefing” and daily recap workflow that functions like an executive assistant, pulling context from connected tools such as Outlook/365, Teams, Monday.com or Notion, SharePoint, and a ticketing system. Kyle describes using Claude Code with simple slash commands (e.g., /morning and /daily recap) to surface meetings, open tasks, priorities, and prep context, then updating to-dos throughout the day to avoid relying on an inbox as a task list. They demonstrate Claude’s integrations, scheduled briefings, and “skills,” clarify skills vs. agents (and token/cost benefits), and discuss persistent memory via tools like Monday/Notion. They share a universal setup guide and prompt template, touch on security/governance cautions, and briefly explain OpenClaw and its risks.

00:00 Morning Briefing Intro

00:30 Building an AI Assistant

01:30 Morning and Daily Recap

03:44 Inbox Overload Problem

05:16 Live Demo Setup

05:29 Connecting Work Tools

07:03 Skills Explained

09:59 Agents vs Skills

14:18 Morning Briefing Walkthrough

15:17 Persistent Memory Systems

17:29 Email and Schedule Highlights

18:53 Tickets and File Examples

19:36 Manual vs Auto Memory

24:40 Choosing Notion or Monday

30:37 Make It Exist Then Good

30:57 Next Skill Idea Triaging

31:45 Cited AI Summaries

32:22 Universal Setup Guide

33:02 Tool Stack Options

33:43 Add Personal Sparkle

35:48 Data Governance PSA

38:25 Setup Paths and Automation

39:03 Cowork and Agent Actions

40:38 Copilot and Platform Roadmap

47:13 Prompt Template Walkthrough

50:00 OpenClaw Explained

53:36 Sandboxing and Safety

58:13 Wrap Up and Resources

Kelsey talks with CIT President and CEO Kyle Etter about recent AI releases like GPT 5.4’s native computer control, Gemini 3.1 updates, and the shift from chatbots to agentic AI that can navigate business systems and complete real work. They discuss why IT infrastructure must now include AI governance, identity, logging, permissions, and emerging liability and insurance considerations for “digital teammates,” noting early examples like insured voice agents. Kyle advises deploying narrow, task-specific agents with strong guardrails, human-in-the-loop checks, and layered QA/security agents to reduce risk. They also cover preventing shadow AI by providing authorized business licenses, training staff to manage agents, and why ROI is increasingly driven by integrating agents into existing software and data workflows, including analysis and data cleanup.

00:00 Welcome and AI Shift

00:50 GPT 5.4 Computer Control

03:16 Governance and Liability

06:31 Designing Digital Workers

12:05 Risks and Guardrails

18:35 Jobs and Agent Managers

23:34 Stopping Shadow AI

26:04 Training and First Use Cases

28:51 AI Licensing ROI

29:35 Software Over Hardware

30:26 Agents Deliver Value

33:57 Data Cleanup With AI

36:58 Conversational Analytics

40:34 Automations Built Faster

43:34 Governance And Guardrails

44:27 Capture Ideas With Tools

48:29 AI To Do Assistant

52:03 Wrap Up And Resources

In this third LinkedIn Live AI chat, Kelsey and Todd discuss how to define and prove AI ROI, arguing many AI projects fail due to poorly defined returns beyond simple time-savings. They distinguish AI activity from AI value, stressing goal-driven, iterative adoption and measurable milestones executives and CFOs can trust—typically cost savings (growth without headcount) or revenue impact (more leads, faster launches, higher win rates). They walk through ROI approaches for practical scenarios including SOP creation and searchable copilots, marketing email HTML generation, new-hire self-serve process help, AI-assisted RFP drafting, job descriptions and onboarding templates, automated revenue-report narratives, customer success call recaps, contract review prep, and meeting-to-action-item pipelines, emphasizing reduced rework, fewer dropped tasks, better decisions, improved customer satisfaction and retention, and lowered legal and operational costs.

00:00 Live Chat Kickoff

00:14 Defining AI ROI

01:39 Activity Versus Value

02:33 Proving ROI to CFOs

04:19 Measuring After the Fact

06:04 ROI Scenario Challenge

06:42 SOPs and Process Consistency

09:36 Marketing Email Automation

10:49 Onboarding and New Hire Support

12:53 Sales RFP Drafting

13:58 HR Docs and Consistency

15:44 Finance Reporting Summaries

18:55 Customer Success Recaps

20:27 Legal Contract Review

22:06 Meeting Actions and Accountability

24:00 Todd’s ROI Tips

26:02 Wrap Up and Next Episode

In this LinkedIn Live episode, host Kelsey and Todd discuss how to improve AI workflow results by moving beyond vague prompts like “write me a marketing email” and using an iterative approach. Todd shares the CRIT framework (Context, Role, Interview, Task), recommends prompting AI to ask clarifying questions one at a time, and stresses specifying desired outputs such as a Word document, table, or executive summary. They demo a prompt-building tool called Prompt Cowboy and discuss using screenshots or PDFs to extract, simplify, or reformat information, including translating legal documents into plain English. The conversation highlights governance concerns such as AI memory, paid vs. unpaid tools, and avoiding sensitive data like PII, employee information, financials, and health details. Todd also mentions agents for specialized compliance use cases and a voice tool called Whisper Flow for capturing ideas while away from a keyboard.

00:00 Welcome and Setup

00:55 Why Prompts Fail

01:49 CRIT Prompt Framework

04:10 Handling Complex Requests

05:50 Prompt Cowboy Demo

07:58 Iterative Prompting Tips

10:32 Screenshots and Privacy

16:23 Everyday Business Use Cases

19:04 Saved Prompts Debate

26:44 Agents and Compliance

27:53 Wrap Up and Voice Tools

30:05 Outro and Next Episode

Kyle (CEO & President at CIT) and Brett (VP of Technical Sales at Genians) join us to discuss recurring and emerging security challenges, including protecting the inside of the network, visibility into who and what is connected, and meeting audit requirements in regulated industries like credit unions and manufacturing. They discuss how VPN zero-days and compromises render perimeter-only defenses insufficient, emphasizing network access control (NAC), device trust policies, alerts, segmentation, and zero-trust principles such as least privilege. They note many organizations lack accurate asset inventories and are surprised by the number of connected devices, including rogue or unmanaged equipment. For prioritization, they recommend starting with core best practices (firewalls, EDR, MFA) and moving toward NAC and ZTNA-style remote access as VPN risk increases. They also cover budget and complexity barriers, product overlap, analysis paralysis, and practical AI adoption examples like Slack AI and Microsoft Copilot for search.

Learn more about Genians: https://www.citsolutions.net/about-computer-integration-technologies/partners/genians/

00:00 Security Challenges Today

01:18 Inside Network Threats

03:14 VPN Zero Days Defense

04:13 Why Problems Persist

08:40 Prioritizing Security Steps

10:50 Budget and Complexity Barriers

13:34 Hidden Devices Discovery

18:05 Common Missteps and Overlap

19:48 Practical AI Use Cases

24:08 Action Steps and Wrap Up

In the first AI Live chat on LinkedIn, Kelsey interviews CIT President and CEO Kyle about how he uses Claude Code as an agent-based executive assistant running locally on his MacBook. Kyle describes building MCP connectors into business systems like Halo (ticketing/PSA), NetSuite (financials), email, and team chats, plus using “skills” that load only when needed to save tokens and cost. He explains how he prompts in plain language while the AI helps build the setup, and shares examples like generating daily “morning” and end-of-day recaps, creating customer health summaries, and quickly producing spreadsheets to identify inactive customers by correlating invoicing and support data. They discuss rapid shifts in AI-assisted coding, data quality checks, ROI, and how AI may transform roles more than replace them.

00:00 Welcome And Setup

00:47 CEO Nighttime AI Habits

01:45 Building A Local Agent

03:44 Daily Workflow With Claude

05:11 Plain Language Development

07:37 Claude For Everyone

08:54 Skills And MCP Explained

14:19 AI Writing Software Shift

16:59 Productivity ROI Examples

19:56 Quick Hit Use Cases

20:55 Voice Notes to Tasks

21:19 Instant Customer Checkins

23:14 Reengaging Stale Accounts

26:21 Access Permissions and Data Quality

28:07 Validating AI Outputs

29:28 Cost Value and Key Users

31:44 Daily Recap Replacing Dashboards

33:37 Agents and the Future of Work

36:05 Replacing vs Transforming Roles

41:35 Adoption Advice and Wrap Up

Human Oversight, Transparency & What It Means for Global AI Regulation

In this episode, Todd (COO & CISO) and Nate (Director of Cybersecurity) to discuss South Korea’s new AI Basic Act and what it signals for AI regulation globally. They outline the law’s focus on guardrails such as mandatory disclosure when AI is used, clear labeling of AI-generated content, human oversight in high-impact areas (like healthcare and critical infrastructure), and explainability around how AI-driven decisions are made. The conversation compares South Korea’s approach as a middle ground between the EU’s stricter governance (EU AI Act) and the US’s current emphasis on innovation and some AI deregulation, while noting the US has limited election-related rules around deepfakes. They discuss real-world risks like AI errors in allergen data, misleading or harmful AI content, and business impacts including compliance pressure, the importance of human validation, and maintaining an inventory of where AI is used to address “shadow AI.” They also note that South Korea’s fines appear relatively small compared to EU-style penalties, and that broader impact will depend on enforcement and how regulations evolve over the next several years.

00:00 AI Regulation Kickoff

00:26 South Korea AI Act Basics

01:17 Transparency and Labeling Rules

02:22 Elections and Deepfakes

03:21 Real World Risk Examples

04:39 Global Approaches EU US Korea

05:41 How Global Rules Shift

07:20 US Policy Tensions and Lawsuits

09:00 Business Compliance and Fines

10:10 Explainable AI in Banking

11:16 Regulated Marketing Disclosures

13:23 Shadow AI and Inventories

14:33 Explainable AI Expectations

14:46 Disclosure and AI Skepticism

15:50 AI in Customer Delivery

16:29 Code Risk and Accountability

17:18 Hallucinations in Compliance

18:03 Human in the Loop

18:59 Leadership and Validation

19:50 High Stakes Decisions

20:59 Regulation and Enforcement

21:34 Wrap Up and Next Episode

CEO & President Kyle and Graphic Designer & Brand strategist Kelsey explore how prompting has evolved from using AI like a “smarter Google” to structured strategies that deliver sharper, less generic results.

They break down the CRIT framework (Context, Role, Interview, Task), share why detailed context reduces hallucinations, and explain how prompt libraries and model memory speed up repeatable work. The conversation also dives into context engineering with tools like Microsoft 365 Copilot and Google Workspace Gemini to make AI outputs more relevant and secure.

Plus: common prompting mistakes, model comparisons, multimodal inputs, and how to onboard teams without losing brand consistency.

Listen now to level up how you work with AI.

00:00 Prompting Then vs Now: From “Smarter Google” to Strategic Skill

00:39 Why AI Sounds Vanilla: Averages, Models & AI Slop

01:33 Prompt Engineering & the CRIT Framework

02:35 Interview-Style Prompts: Fewer Hallucinations, Better Results

04:10 Garbage In, Garbage Out: Treat AI Like a New Hire

05:04 Let AI Help Write Prompts + Tools & Libraries

07:08 Why One-Liners Fall Flat (Contractor Analogy)

07:55 From Prompts to Systems: Templates & Model Memory

11:21 Context Engineering: Files, Memory & Workplace Data (Copilot/Gemini)

13:27 Over-Prompting: Context Limits & When to Reset

16:26 Set Outcomes, Don’t Micromanage

18:22 Smarter Models: Gemini & Claude Need Less Steering

19:06 Claude Opus vs ChatGPT: Speed vs Detail

20:27 Multi-Model Workflow: Use Each for Its Strength

21:20 Why New Models Feel Smarter

22:11 Ask AI to Improve Your Prompts

24:42 Planning Mode: Structured Builds & AI Interviews

26:13 Training Teams: Frameworks, SOPs & Safe Experimentation

31:47 Multimodal & Voice Prompting (Gemini’s Edge)

33:15 Wrap-Up & What’s Next

In this episode, we are joined by Cole, a Power Platform Developer, and Kelsey, a Graphic Designer and Brand Strategist, to discuss the key differences between various AI models. The conversation emphasizes the importance of trial and error, context, and the specific needs of different departments within a business. We also touch on the practical applications of GPT, Copilot, and other tools, and explore how businesses can better align their AI strategies. Tune in for a deep dive into AI utilities, effective prompting, and strategic decision-making to boost your business productivity and efficiency.

00:00 Introduction to AI Models

00:27 Differences Between AI Models

01:14 Practical Applications and Preferences

03:05 AI Tools and Their Use Cases

04:22 Choosing the Right AI Tool

08:20 Memory and Context in AI

15:24 Effective Prompting and Tone

20:10 Building Context in AI Conversations

22:26 Balancing Business Tools and Employee Preferences

23:40 Understanding the Gap Between Business and Personal Tool Use

24:20 Tailoring Tools to Department Needs

25:55 The Importance of Pilot Programs for AI Tools

28:48 Governance and Safe Use of AI Tools

30:58 The Creative Process of Using AI Tools

34:22 Effective Communication with AI Models

37:04 Conclusion and Future Discussions