Podcast Summary:

AI Explored — "AI Operating Systems: The Next Level of AI for Business"
Host: Michael Stelzner
Guest: Dr. Nikki Sweeney, Founder, AI Her Way
Date: September 9, 2025

Overview

This episode dives deep into how businesses of all sizes can construct their own AI operating systems—a layered, strategic approach to embedding AI throughout an organization. Dr. Nikki Sweeney shares her journey into AI, explains the five levels of AI maturity for businesses, and offers practical steps to build sustainable and ethical AI frameworks. Whether you’re a marketer, entrepreneur, or business leader, this episode uncovers actionable strategies to leverage AI for efficiency, impact, and competitive advantage.

Key Discussion Points & Insights

Dr. Nikki Sweeney’s Journey Into AI (03:45-06:42)

• Background: PhD in Conservation Ecology; career in data science and higher education.
• Transition to AI: Coding proficiency via PhD requirements; generative AI arrival in 2022 was transformative.
• Mission:
  • Realized the gender gap in AI usage/design (noting 70% male/30% female ChatGPT usage at the time).
  • Founded “AI Her Way” to ensure ethical, diverse, and impactful AI development and adoption.
• Quote:

  "I literally jumped on [my partner] and shook him by the shoulders and I was like, this is what I'm gonna teach people, because this is going to change everything."
  — Nikki Sweeney (05:11)

Why Ethical & Sustainable AI Matters for Business (06:42-09:39)

• Competitive Edge: The real advantage comes not just from using AI, but from using it responsibly.
• Brand Risk: Unethical AI can seriously damage reputation (e.g., Amazon’s biased hiring algorithm).
• Broader Impact: Businesses now shape AI norms for future generations.
• Quote:

  "Just using [AI] is not your competitive advantage... using it in a way that contributes to a greater mission... that is a key differentiator."
  — Nikki Sweeney (07:31)

The Five Levels of AI Maturity in Business (09:58-17:24)

Level 1: Prompting with Large Language Models (LLMs) (10:14)

• Basic engagement: Using tools like ChatGPT or Claude for productivity.
• Skill everyone needs, but just the starting point.

Level 2: Automations & Custom GPTs (11:27)

• Custom bots (“custom GPTs,” “Claude Projects,” “Gems in Gemini”) automate single, repeatable tasks.
• Example: An assistant trained to write discovery call scripts.

Level 3: AI Agents & Assistants (12:37)

• Broader access: Connect to tools like email, calendars, CRM, etc.
• Assign more complex instructions akin to a human assistant.

Level 4: AI Workflows (13:35)

• Tying multiple automations and agents into multi-step processes.
• Example: Automated pre-meeting research, script generation, and email distribution.

Level 5: AI Operating Systems (17:24)

• Entire divisions governed by AI—overarching objectives with autonomous decision-making and delegation.
• Example: An AI ‘Head of Marketing’ orchestrates all campaign tasks and analyzes outcomes.

Understanding the Differences

• Custom GPTs vs. Workflows:
  Custom GPTs automate single tasks; workflows combine multiple steps, often requiring different AI “personas” (14:04-15:18).
• Agents vs. Workflows:
  Agents manage ongoing, semi-autonomous tasks; workflows are predefined chains of multiple automations.

"An AI operating system is like the head of a division in a business... It is the boss of everyone underneath it to execute on your objective."
— Nikki Sweeney (18:37)

Building Your Own AI Operating System: Where to Start (19:13-24:13)

1. Define Objectives & Roles:

   • Identify clear business goals.
   • Design AI “positions” as if hiring staff: what is their responsibility, expectation, department?

2. Example — Marketing AI OS:

   • AI Head of Marketing receives objectives, decides campaign strategy, delegates to sub-agents for social, email, PR, etc.
   • Tracking: All deliverables and progress are monitored via dashboards.

3. Human Integration:

   • AI enhances human roles, freeing people to innovate rather than replace them.
   • Everyone can become more strategic—AI takes on repetitive/process-driven work.

“What AI does is it really flattens organizational structures... you unlock the capability of everyone to sit in a strategic, creative innovation space.”
— Nikki Sweeney (25:00)

Practical Steps & Tools for Building AI Systems (30:32-42:11)

Designing & Documenting Departments

• Top Down: Map out departments, roles, responsibilities, and desired outcomes.
• Visualization Tip: Use AI tools like Claude Artifacts to create system diagrams/flowcharts from written role descriptions (31:21).

Bottom Up: Start Small & Build

• Pick One Repeatable, Process-driven Task:
  Start with an LLM, craft a solid prompt, iterate until consistent results (32:22).
• Graduate to AI Assistants:
  Move successful prompts to custom GPTs or Claude Projects.
• Automate With Workflows:
  Use platforms like Make.com, n8n, or Zapier to automate processes that mirror real employee workflows. (33:28-35:27).

Knowledge Base: Where Does the Data Live?

• Preferred Tool: Notion (35:50-38:41)
  • Flexible, plays well with automations, integrates Google Workspace.
  • Built-in AI for search and summarization supports both human and AI staff.

Bringing It All Together

• Centralize everything (decisions, outputs, progress) in Notion.
• Use automations and agents to trigger, track, and manage all processes via this central dashboard.

Agents in Action & Workflow Hierarchies (41:56-45:46)

• AI Agents:
  Beyond basic automations, agents can autonomously monitor and act on broader directives, e.g., continuously scanning inboxes for relevant content.
• Security Warning:
  Agents need broad data access; only deploy when confident in your system-building and quality control (42:11-45:16).

“Prompting with a large language model is like having an intern sit right next to you... The AI operating system is like having a whole division that kind of sits in a whole nother office."
— Nikki Sweeney (44:23)

Overcoming Overwhelm & Getting Started (45:46-48:49)

• It’s normal to feel overwhelmed.
• Start with ONE thing:
  • Choose a repetitive, process-driven task you can explain clearly—and that doesn’t bring you joy.
  • Write prompts like you’re instructing an eager intern.
• Build upwards:
  • Refine → Automate → Integrate into operating system over time.
• Recommendation: Write down all tasks, mark repetitive/annoying ones, select a starting point.

Notable Quotes & Memorable Moments

• "This is going to change everything... and for me it just kind of ticked all those boxes. You know, it was data, it was code, but it was much more human... I just thought it was so fascinating and such an important space to be in."
  — Nikki Sweeney (05:00)

• "The real market value is doing it in an ethical way... If you don’t do it an ethical way, you stand a real risk of damaging your brand, your business, your reputation—either as an individual or as a company."
  — Nikki Sweeney (07:22)

• “If we don’t do [ethical AI] right now, we miss our chance to contribute to that [future].”
  — Nikki Sweeney (09:26)

• "You want to baby it... It’s really important to take it slowly. You will get there and your end result is going to be way better than if you just jump in at the top."
  — Nikki Sweeney (44:54)

• "Pick something that ticks all four boxes: it’s repetitive, process-driven, you could explain it really well, and it doesn’t fill you with joy."
  — Nikki Sweeney (46:45)

Timestamps for Key Segments

• Nikki's AI Journey: 03:45-06:42
• Ethical/Sustainable AI for Business: 06:42-09:39
• 5 Levels of AI for Business: 09:58-17:24
  • Custom GPTs vs Workflows: 14:04-15:18
  • Agents vs Workflows: 15:18-16:17
• Defining & Building an AI OS: 19:13-24:13
• AI for Human-Driven Teams: 24:13-27:02
• Practical System Design Steps: 30:32-34:13
• Automations & Knowledge Bases: 34:13-41:56
• Agents vs. Automations: 41:56-45:46
• Dealing with Overwhelm: 45:46-48:49

Resources

• Dr. Nikki Sweeney — AI Her Way:

  • Website: www.aiherway.com
  • Instagram: @aiherway
  • LinkedIn: Dr. Nikki Sweeney
  • Special Resource Page: pages.aiherway.com/SME (includes video walkthroughs and blueprints)

• Knowledge Base Tool: Notion

• Automation Tools: Make.com, n8n, Zapier

• AI Assistants: ChatGPT, Claude, Gemini, Perplexity

Final Takeaway

Starting your AI journey doesn’t require replacing people—it’s about maximizing both human and machine strengths in harmony. Start with manageable, well-defined tasks and gradually build your way up to a full AI operating system. The key advantages come from ethical, thoughtful design and maintaining a collaborative human-AI environment.