Unlocking ROI with AI in Chemicals | Insights from Kendall Justiniano - Ep. 267

The Chemical Show: Where Leaders Talk Business32mJune 9, 2026
AI-Generated Summary

The exploratory phase of AI in the chemical industry is over, and 2026 is the year of return on investment—according to Kendall Justiniano, managing director of Growth Arc Advisors. In a candid conversation with host Victoria Meyer, Justiniano argues that most companies are still stuck in the early stages of AI adoption, relying on chat prompts and isolated tools like Copilot, which deliver minimal ROI. The real value emerges only when organizations move beyond chat to connect AI to enterprise data systems—CRM, ERP, and document repositories—through what he calls 'cloud code' and AI agents that operate across workflows. He describes a future where AI lives as a contextual, architecture-layered interface that mimics human cross-system navigation, automating repetitive tasks in procurement, sales, and marketing. The key to success? Governance, internal AI expertise, and embedding organizational knowledge into secure, proprietary context layers that become competitive advantages. Justiniano shares that his firm now runs fully AI-enabled engagements, with automation from concept to deployment, and warns that without proper architecture and validation, even advanced AI systems fail. The future isn’t about replacing humans—it’s about augmenting them with systems that understand the business context at scale.

Key Takeaways
1

Move beyond chat prompts: AI ROI requires connecting to enterprise data systems like CRM and ERP via AI agents, not just using chat interfaces.

2

Build proprietary context layers: Embed organizational knowledge into AI systems to create defensible competitive advantages that can’t be replicated by vendors.

3

Governance is non-negotiable: Unbounded AI access creates risk; executives must establish controlled, expert-led frameworks for AI deployment.

4

AI excels at transforming unstructured data: Language models are ideal for processing contracts, invoices, and emails—common in procurement and sales workflows.

5

Automate workflows, not just tasks: The highest ROI comes from chaining AI actions across systems (e.g., pulling MSDS sheets, applying selection guides, pre-selecting products).

…and 3 more takeaways available in PodZeus

Chapters
0:00
2 min

Welcome to The Chemical Show: AI’s Transition from Exploration to ROI

Victoria Meyer introduces the episode and guest Kendall Justiniano, setting the stage for a discussion on AI’s evolution in the chemical industry, marking 2026 as the year of return on investment after a prolonged exploratory phase.

2:00
3 min

Kendall’s AI Journey: From Chat to Cloud Code and Full Integration

I went deep into the other end of the pool... and frankly, it was scary when I did it because I'm running Claude code at the terminal line and I'm not a coder.

Highlight
5:00
3 min

The Myth of Grassroots AI Adoption: Why Chat Prompts Don’t Scale

As long as you're at the chat prompt, you're not discovering. You're not going to discover through a grassroots mechanism where the interesting opportunities are for AI.

Highlight
8:20
3 min

From Chat to Cloud Code: The Power of Connected Data Systems

When you start giving people access like that, or when you start experimenting with that type of access, now you're working at scale. Now you're working with the entire breadth of the data that you have available to you.

Highlight
11:40
3 min

AI as the ‘Octopus Layer’: Architecture, Governance, and Human-Centric Design

The AI is going to live close to your organization. It's not embedded in a system, it's going to live next to your organization and if it's doing work that people was too time consuming but it's still human work.

Highlight
High-Impact Quotes
In other words, the context is provided and it's provided independent of the operating model. So literally, I can swap out Cloud Code and go grab another LLM and say, read my specification, and it will know how to operate within the environment that I've provided that specification for.
Kendall Justiniano30:08
So that was a transformational moment when I dove into the other end of the pool. And frankly, it was scary when I did it because I'm running Claude code at the terminal line and I'm not a coder.
Kendall Justiniano5:35
We don't trust the AI but if you run those systems at scale with appropriate contexting and appropriate protection of that context, hallucination isn't the kind of issue that it is or that people see in fear when they're sitting at that chat prompt.
Kendall Justiniano26:28
Speakers

Host

Victoria Meyer

Guest

Kendall Justiniano
Topics Discussed
ai in chemicals95%ai roi90%enterprise ai integration88%context layers86%ai governance85%llm applications82%ai workflow automation80%data connectivity78%
People & Brands

Kendall Justiniano

person

18xPositive

Growth Arc Advisors

organization

12xPositive

Claude

product

8xPositive

Copilot

product

5xNeutral

CRM

other

4xNeutral

Cast Magic

product

4xPositive

ERP

other

4xNeutral

ChatGPT

product

3xNeutral

McKinsey

organization

2xNeutral

Booz & Co

organization

2xNeutral

Start discovering podcast insights today

Start with a 7-day trial and explore a growing catalog of popular podcasts. No credit card required.

No credit card required • 7-day trial • Cancel anytime