Product Manager Script (30 Minutes)
Thanking & Transitioning from VP (~2 minutes)
Thank you, [VP’s Name].
I truly appreciate how you’ve highlighted that AI is already part of our everyday lives, clarified
the essentials of AI and Generative AI, and connected them to our company’s strategy,
values, and people-first approach.
we’re running across the company.
R E
That context is invaluable, and I’m grateful to now build on it by sharing some of the initiatives
U TU
Good evening, everyone.
| F
Greeting & Introduction (~3 minutes)
O N
T I
I feel truly grateful to be here with all of you today, and I’m excited to continue this insightful
conversation together.
V A
Before we dive deeper, let me introduce myself. I’m [Your Name], an AI Product Manager
NOhere at [Company Name]. My role is about helping the business use AI and Generative AI
where they can add the most value.
I N I work closely with our technology teams to bring solutions into practice, and I also support
multiple business functions in adopting RPA and automation.
In short, I focus on making sure we adapt the right technologies, in the right way, at the right
time.
Setting Expectations (~2 minutes)
We’ve just explored how AI connects to our company’s values and why keeping human in the
loop remains central to our approach.
Now, I’d like to focus on two key pilot programs that are putting these ideas into practice:
Project Creed, and MyCompany GPT.
Together, they highlight how we’re moving from awareness into action, testing real use cases
that can bring meaningful value across our business.
Transition into Project Creed (~n/a)
To bring our AI strategy to life, we needed to start small, test carefully, and solve a problem
that truly mattered to our people. That’s how Project Creed began — as our first pilot
designed to make life easier for our call center associates.
Project Creed (~7 minutes)
R E
TU
The challenge we set out to solve was clear.
F U
Our call center associates often need to search through more than 6,500 knowledge-based
articles in Genius. That process can be slow, leading to delays, escalations, and customer
frustration. |
O N
T I
In fact, data showed that in over half of cases, Genius wasn’t being used before calls were
escalated to CARES — that’s around 50,000 calls every single month.
V A
That gap was significant, and that’s exactly where Project Creed comes in.
NO
I N As part of Project Creed, we’ve been testing a Generative AI–powered chatbot designed to
assist associates.
Our Incubation team built a minimum viable product and tested it with a small group of users.
The chatbot was built on AWS Bedrock, with about 85 Genius documents loaded from the
Collections space.
It could even remember conversations, allowing associates to ask follow-up questions
naturally.
Here’s how it works: an associate types a question in plain language. The system then
searches Genius for relevant information. Generative AI provides a clear, concise summary.
Associates can ask follow-up questions to refine the answer. And importantly, all responses
are linked back to the original source for verification.
The potential benefits of Project Creed include shorter customer contact time, improved
experiences, more accurate information, productivity gains, and faster proficiency for new
associates.
Transition into MyCompany GPT (~n/a)
Project Creed taught us how generative AI could support associates in the moment of need.
But we also realized there was a bigger opportunity — to give every employee a secure,
company-specific AI assistant they could rely on every day. And that’s exactly what we set out
to create with MyCompany GPT.
MyCompany GPT (~7 minutes)
R E
TU
MyCompany GPT is a custom-built interactive AI, tailored for our company’s data, policies,
and security requirements.
F U
Unlike public tools such as ChatGPT, company data remains fully private. That means we get
|
the benefits of generative AI with the assurance that it’s secure and aligned with our
standards.
O N
T I
You can use MyCompany GPT to create or summarize content and research
A
information-heavy topics.
V
NOWhy use it? It automates repetitive tasks, saves time, provides quick access to knowledge,
I N and frees us to focus on meaningful work.
Its capabilities include secure login with Okta, support for public, internal, and confidential
data, conversation management and history, exporting conversations, prompt shortcuts,
multiple LLM options, and even a beta file upload feature with enhancements on the way.
Now, when using MyCompany GPT, there are a few best practices. First, specify — add
context, assign GPT a role, and define the audience. Second, clarify — stick to one question
and break down big asks. Third, review and edit — fact-check, consider compliance, and
follow processes.
For example, instead of asking, ‘I work for MyCompany. Tell me about the risks in our IT
system,’ you could say, ‘Act as an Information Security Officer. Provide a detailed assessment
of cybersecurity risks faced by financial institutions over the past six months, focusing on data
breaches and unauthorized access.’
Another example: instead of asking, ‘How can we reduce IT risks?’ you might say, ‘What are
the most effective risk mitigation strategies for minimizing third-party vendor risks caused by
data leakage incidents in financial institutions such as retail banking or credit card providers?’
And with that, I’m happy to share that MyCompany GPT is now available to all exempt
employees.
Transition into Demo (~1 minute)
We’ve covered what MyCompany GPT is and how to use it effectively — now I want to show
E
you how one well-structured prompt can transform into a full knowledge article.
R
MyCompany GPT Demo (~5 minutes)
U TU
| F
In this demo, we asked MyCompany GPT to create a structured knowledge base article on
paper statement fees and related policies.
O N
T I
The key was not just to ask a question, but to give very specific instructions: who the audience
is, what format the article should take, and what outcome we need.
V A
By doing that, the AI produced content that was accurate, policy-aligned, and easy for
NOassociates to use in real customer conversations.
I N This demo shows how, with the right prompts, MyCompany GPT can go beyond quick
answers and actually generate full, usable knowledge that saves time, reduces errors, and
ensures consistency.
Closing Recap (~3 minutes)
So we’ve now discussed both of our pilots — Project Creed and MyCompany GPT — and
looked at how they’re starting to shape the way we work.
At a high level, today’s session has been about three things:
First, understanding what AI and generative AI mean for us.
Second, connecting that understanding to our company’s values and strategy.
And third, seeing how pilots like Creed and GPT are turning those ideas into action.
I’m truly grateful for the opportunity to share this with all of you.
I hope this session has sparked new ways of thinking about AI, generative AI, and the
opportunities they can bring to your work and our organization.
That brings us to the end of our session. Now, [Other Presenter’s Name] and I would be glad
to hear any reflections, clarifications, or points you’d like us to expand on.
R E
U TU
| F
O N
T I
V A
NO
I N