Podcast Summary: The Agile Brand with Greg Kihlström® – Episode #743

Title: Always-on collaboration between AI and marketing with Devin Knysz, Salesforce
Release Date: October 1, 2025
Guest: Devin Knysz, Senior Director, Industry GTM Consumer Goods and Retail, Salesforce
Host: Greg Kihlström

Overview

This episode explores the always-on collaboration between autonomous AI agents and marketing teams within enterprise organizations, focusing on Salesforce's unique agentic AI approach ("Agent Force"). Through practical examples and expert insights, the conversation highlights how artificial intelligence is transforming customer experience (CX), sales, and internal workflows by automating routine tasks, freeing up employee time for value-added activities, and creating new opportunities for personalized engagement.

Key Discussion Points & Insights

1. What is Agentic AI?

[03:20 – 04:29]

• Definition & Salesforce’s Position:
  Agentic AI refers to autonomous AI agents acting as continuous collaborators, handling complex tasks based on actionable data.
• Salesforce’s Unique Approach:
  • Agent Force is "grounded where your humans do their work;” it's embedded within the same consoles (marketing, service) staff already use.
  • Deep integration with Salesforce’s CRM provides rich context, outpacing mere LLM (large language model) point-solution agents.
• Quote:

  “Our Agent Force product is also grounded where your humans do their work...in the application layer where your humans do their work.”
  — Devin Knysz [03:32]

2. Case Studies in Agentic Workflow

[04:38 – 07:17]

• Consumer Goods Example:
  • For a home hardware firm with frequent faulty part complaints, Agent Force automates the entire warranty process: identifying the caller, matching parts, automating shipment, and enriching the customer profile—all without human involvement for ~60-80% of cases.
  • Impact: Accelerated resolutions, improved CX ("my faucet is definitely going to break at 7:30 at night... not when your service center is open”), and more meaningful human work.
• Human Value Elevation:
  AI enables employees to shift from repetitive tasks to “advisors”—as seen in the Ikea reskilling example.

3. The Future of Human Roles in Sales & Service

[07:42 – 10:51]

• Shifting Human Value:
  • AI agents handle admin—appointment setting, information retrieval, objection management—so sales reps and CX staff focus on strategic human interaction.
  • Real-world application: Agents can instantly supply detailed product info (e.g., product numbers, specs), giving reps more time for high-impact work and personal life.
• Quote:

  “The appointment setting is not the value that I have...there’s more work than anyone can do in a day...this is allowing you to free up that work...”
  — Devin Knysz [08:14]

4. Pre-built AI Skills & Guardrails

[12:03 – 15:05]

• Pre-built Skills:
  Salesforce offers libraries of “skills” (templates for tasks like proactive maintenance, sales agreements) so organizations can deploy agents rapidly—weeks, not months.
  • Configuration remains necessary for company-specific workflows.
• AI & Data Trust:
  • Salesforce prioritizes trust: “None of the data that goes through our platform is used to train models.”
  • Human-in-the-loop remains a key safeguard: humans approve non-trivial automated actions.
• Notable Moment:
  • Cautionary tale of ChatGPT “leaking” RFP competitive info after being trained on customer data—reinforces need for trusted platforms.

5. Language as the New User Interface

[15:26 – 16:05]

• The paradigm shift: “It’s kind of cool to be in today’s day and age where language becomes the primary interface.”
  • Tools like drag-and-drop, natural language bots now allow non-coders to prototype powerful solutions.

6. Retail & Field Agent Transformation

[16:32 – 20:10]

• Retail Visit Management:
  • AI agents now handle scheduling, briefings, and data gathering, surfacing top recommendations for store visits in real time.
  • Formerly generic “A/B/C store” routines are now replaced by context-rich, zip code–specific recommendations, streamlining short in-store windows for reps.
  • Lower-volume locations that rarely receive in-person visits can now be serviced by digital agents, closing CX gaps.
• Impact:
  More efficient resource use, higher personalization, and better CRM enrichment.

7. Agentic AI & Customer Service Reps With Quotas

[20:10 – 23:44]

• Visionary Application:
  • Beyond basics like “where is my order?”, agentic AI can handle nuanced support (e.g., finding and shipping a lost bolt for a lawnmower), making processes smooth for customers who avoid 1-800 calls.
  • Companies gain richer end-user data and improve the B2B consumerization trend.
• Quote:

  “Anybody under the age of...let’s just call it directionally 40, where like anybody under that’s not calling an 800 number.”
  — Devin Knysz [23:22]

8. Where Should CMOs Start?

[23:44 – 28:00]

• Actionable Advice:

  • “Pick a real problem” or pain point to apply agentic AI. Don’t try to boil the ocean—start with achievable pilots (e.g., website experience, personal shopper) and build from there.
  • The role of brand websites is shifting: as AI reduces the number of clicks, brands must rethink how their digital properties engage and convert customers, potentially by creating “branded AI experiences.”

• Quote:

  “I think the most important element...is pick a real problem.”
  — Devin Knysz [24:15]

• Experimentation Encouraged:
  Utilize vendor expertise—Salesforce, for instance, has deployed over 5,000 agentic AI agents.

9. Staying Agile in AI

[28:00 – 29:00]

• Devin’s Practices:

  • Tries every new AI tool.
  • Sets reminders to revisit tools as their capabilities evolve—“Every time you experiment and it can’t work, three months later it can.”
  • Recent discovery: Gemini can auto-update his calendar from emails, saving time and freeing him up for further exploration.

• Quote:

  “That’s my best advice: push the limits of AI because it’s such a rapidly expanding space.”
  — Devin Knysz [28:53]

Notable Quotes & Timestamps

• On Agentic AI and Human Value:

  “We are grounded in not only the workflow that your humans do, but also all of the rich CRM data that comes with that.”
  — Devin Knysz [03:35]

• On Automating Top Complaint Claims:

  “We're seeing people handling 60, 70, even 80% of interactions autonomously, which in my opinion frees up humans to do more interesting work.”
  — Devin Knysz [05:56]

• Sales Use Case:

  “If you’re selling anything the odds are your bag is too big for you to memorize all the key things.”
  — Devin Knysz [09:52]

• On Human-in-the-loop Guardrails:

  “It’s more about humans becoming reviewers and approvers versus doers. That’s my other guardrail.”
  — Devin Knysz [14:37]

• On New Role of Language in Tech:

  “It’s fun to watch language and your natural language become the driver of your interaction with software and technology.”
  — Devin Knysz [15:54]

• On Changing Website Roles in AI Era:

  “I think...the role of the website is going to change pretty dramatically in the post-AI world because our funnel has essentially blown up.”
  — Devin Knysz [24:32]

Segment Timestamps

• 00:10 – 01:51: Intro and guest welcome
• 03:20 – 04:29: Salesforce’s definition and approach to Agentic AI
• 04:38 – 07:17: Service center case study and impact
• 07:42 – 10:51: Sales, human role elevation, and live product info
• 12:03 – 15:05: Prebuilt AI skills, configuration vs. customization, and guardrails
• 15:26 – 16:05: Language as the primary interface
• 16:32 – 20:10: Retail visit agentic transformation
• 20:10 – 23:44: Visionary future of agentic customer service
• 23:44 – 28:00: Advice for CMOs and how to start; changing role of websites
• 28:00 – 29:00: Devin’s approach to staying agile with AI tools

Tone & Style

Devin Knysz is practical yet visionary, weaving examples from both his operator and technology leadership backgrounds. The dialogue remains approachable for tech-savvy leaders but is rich with insights for those actively exploring or deploying AI-driven CX and marketing solutions.

Summary

This episode delivers an in-depth look at agentic AI—where agents act as persistent, context-aware collaborators, relieving humans of tedious tasks and elevating the human role to value-added strategy and creativity. With enterprise-ready frameworks, prebuilt skills, and robust guardrails, Salesforce’s Agent Force shows how large organizations can adopt AI for immediate, tangible CX and operational gains—reminding leaders to start small but think big about AI’s transformative potential.