The Agile Brand with Greg Kihlström®

Episode #787: Tray.ai's Stephen Stouffer on AI Adoption and Why Marketers Should Care about the Integration Layer
Date: December 22, 2025
Guest: Stephen Stouffer, Director of Automation Solutions, Tray.ai
Host: Greg Kihlström

Episode Overview

This episode focuses on the critical, often-overlooked importance of the integration layer in enterprise AI adoption. While many companies invest heavily in AI models and applications, they encounter severe roadblocks due to legacy infrastructure and disconnected systems. Stephen Stouffer of Tray.ai unpacks why marketers should focus on integration-first strategies to scale AI and automation, drive ROI, and build organizational agility.

Key Discussion Points & Insights

1. The Integration Layer: The Hidden Bottleneck for AI Adoption

• Host Frames the Challenge:
  • “Why are so many companies spending billions on AI only to see their most ambitious projects stall out before they ever impact the customer experience?” (01:20)
  • Real transformation requires the ability to connect disparate systems and data for meaningful AI at scale.
• Stephen Identifies the Blind Spot:
  • The initial AI use case typically works well as point solutions target them directly. Trouble comes with the second, third, and fourth use cases when “a neighboring department…wants to do it, but their tech stack is just a little bit different,” causing platforms to “get promoted into incompetency.” (04:49)
  • Many platforms lock teams into ecosystems, making wide-scale integration a massive challenge.

2. What Is Tray.ai and What Problems Does It Solve?

• Platform Overview:
  • Tray.ai is a “workflow orchestration and an AI orchestration platform,” enabling companies to build, automate, and scale agents on top of any existing tech stack. (03:14)
  • It is platform-agnostic: “We don’t care whether you have a Salesforce-first approach for your CRM or if you work with Eloqua…Everything is APIs, and we are the orchestration layer that sits on top.” (04:49)

3. The Value of “Integration-First” Mindsets

• Strategic Importance:
  • Integration must be foundational, not an afterthought. Otherwise, even good data remains siloed, and AI agents lack context:
    • “If you not only have bad data, or even if you have really good data, but the data is not connected...the agents are only as good as the data that they have and the connectedness they have within the platform.” (06:25)
• Practical Advice for Marketers & CX Teams:
  • Before investing in AI or automation, ensure your tech stack “at a fundamental layer is connected.” (06:40)

4. Symptoms of Poor Integration in Enterprises

• Organizational Pain Points:
  • IT teams get bogged down with “common issues—password resets, access control” and can’t spend time on strategic projects. (09:06)
  • As use cases spread, disparate teams want to automate similar processes but lack a unified way to do so.

5. Case Study: Yext’s Rapid Integration

• Yext built over 100 integrations in three months using Tray:
  • Their IT created “a series of agents that sat in front of the IT team to do the case deflection”—handling repetitive requests and allowing human staff to focus on higher-value work. (09:03)
  • “You can’t do that with just code. You have to have a scalable infrastructure, which Tray definitely has.” (10:13)

6. Identifying High-Value Use Cases and Selling the Vision Internally

• Start Small, Show ROI, Scale Up:
  • Meet with teams, document repetitive or manual tasks, and target those for automation.
  • Directly tie automations to business goals: “You can directly tie it to revenue…get those real ROI numbers and bring that up to your executive team.” (10:50)
• Metrics to Measure and Communicate ROI:
  • IT: Reduction in ticket open-to-resolution time, # of tickets resolved
    • “How long a ticket is open to when it can be resolved is a key metric...” (12:15)
  • Marketing: Time from lead capture to entering sales pipeline
  • HR: Speed and efficiency of employee onboarding

7. “Future-Ready” Integration: Ensuring Flexibility for What’s Next

• Adapting to Platform Changes and Technology Evolution:
  • Companies must expect change: “We’re using X tool today, we’re probably going to be using Y tool in three years.” (17:12)
  • Tray’s “callable workflows” enable composable automation, so swapping providers (e.g., Clearbit to ZoomInfo for enrichment) requires changing one workflow, not 50. (17:19)
  • “We have that composable architecture—you're only changing the process in one workflow...others don’t even realize under the hood that you've actually changed something.” (18:10)
• Agent Agnosticism:
  • Tray’s agents can run on different foundational models (OpenAI, Gemini, AWS Bedrock) and adapt as new ones emerge.

8. How to Approach Integration-First Transformation

• Collaborative, MVP-Oriented Approach:
  • Marketing execs should ask teams about bottlenecks (“how long an event lead takes to get into sales arms”); if it’s “days,” it’s a problem.
  • Give teams latitude to dig deeper and identify pain points.
  • Build MVPs (Minimal Viable Products) for key use cases, pilot solutions, demonstrate value, then scale. (21:26)

9. Compliance and Data Privacy: A New AI Frontier

• Stephen’s Prediction:
  • “Content on the Internet…it used to be very clearly AI…But it’s getting so good…We’re going to be living in a world…where you’re not going to be able to tell the difference…Controls [for compliance or transparency] need to be put in place.” (22:51)
  • Next wave: standards and compliance for AI-generated content, possibly requiring businesses to disclose AI use or adhere to new certifications.

Notable Quotes & Memorable Moments

• On Point Solutions Becoming Obsolete as Needs Scale:

  • “Point solutions get promoted into incompetency…It started so good, and then it kind of became a problem.”
    —Stephen Stouffer (04:49)

• On Integration Agility:

  • “We are platform-agnostic. We don’t care whether you have a Salesforce-first approach or if you work with Eloqua.”
    —Stephen Stouffer (04:49)

• On Buy-In and Identifying Use Cases:

  • “Meet with your team, document those repetitive tasks…you can get a lot of buy-in when you say, ‘Hey, we can remove that from your plate…build an agent on it.’”
    —Stephen Stouffer (10:39)

• On Measuring Value:

  • “From getting a lead into the system to getting it in the hands of sales took over a week…All that could be automated, done within minutes.”
    —Stephen Stouffer (12:15)

• On Agility and Flexibility:

  • “Callables and composable workflows mean you’re only changing the process one time…others don’t even realize under the hood that you’ve changed something.”
    —Stephen Stouffer (18:10)

• On Staying Agile:

  • “I drink lots and lots of coffee. … I say yes when I want to say no. Things are very scary…have any of us done this before? … So, lots of coffee and say yes instead of no.” —Stephen Stouffer (24:09)

Important Timestamps

• [01:20] – Greg sets up the central challenge: legacy infrastructure stalling AI’s impact
• [03:14] – Stephen explains Tray.ai’s offering and who it serves
• [04:38] – Symptoms of integration blindness in large organizations
• [06:25-07:24] – Garbage in, garbage out: why data connectivity matters
• [09:03] – Yext case study: 100 integrations in 3 months
• [10:39] – Approaching change with buy-in and measurable wins
• [12:15] – Example metrics across IT, Marketing, HR
• [17:12] – The future-ready mindset: anticipating tech/platform shifts
• [18:10] – Callable workflows and composability
• [22:51] – AI compliance and undetectable AI content: the next regulatory focus
• [24:09] – How Stephen stays agile: humor and embracing new challenges

Summary of Actionable Takeaways

• Conduct an Audit: Assess current integration pain points—how long do routine processes actually take?
• Prioritize Integration: Make connectivity and data orchestration the first step, not an afterthought.
• Start Small, Prove Value: Pilot automation on bottleneck tasks, measure time saved, and scale iteratively.
• Choose Platform-Agnostic Solutions: Focus on infrastructure that’ll remain flexible as tools and models evolve.
• Stay Mindful of Compliance: Prepare for heightened regulation around AI and data handling.

This episode provides a practical, experience-backed look at why and how marketers and business leaders must elevate integration and orchestration in their AI and automation roadmaps to unlock real enterprise value.