This Senior Director of Program Management Breaks Down the AI Tools and Integrations She Uses to Bridge Gaps in Her Clients’ Tech
AI Experimentation: AI tools require trial and error, fostering strategic leadership over administrative tasks in project management.
Strategic Leadership: AI reduces busywork, enhancing focus on planning, risk management, and solving broader strategic issues.
Essential Tools: Tools like Elvex and Claude streamline workflows, allowing project managers to be more efficient and effective.
AI Integration Challenges: Integrating AI tools across systems remains difficult due to siloed ecosystems and varied tool performance.
Future Leadership Gap: AI's rise may shift roles, necessitating a focus on leadership and people skills in project management.
Lauren Selley is the Senior Director of Production at Code and Theory, where she oversees the end-to-end delivery of digital solutions across web and mobile. With AI now embedded in her workflows, she’s spending less time on administrative tasks — and more time driving strategic decisions.
We spoke to Lauren about what’s changing in the delivery world, what tools are making a real impact, and how she’s helping her team navigate the shift from project manager to strategic leader. Here's what she had to say.
Why Program Management Requires AI Experimentation
I’m a Senior Director of Production (Program Management) at Code and Theory.
Unlock for Free
Create a free account to finish this piece and join a community of forward-thinking leaders unlocking tools, playbooks, and insights for thriving in the age of AI.
Already a member? Log In
I oversee the end-to-end delivery of digital solutions for our clients — from strategy through launch and post-launch support across web and mobile. I spend a lot of time making sure our teams have what they need to deliver successfully, and that our clients feel confident in us as partners.
In our business, AI is all about experimentation: test, optimize, repeat.
Companies are not handing us polished solutions that say, “Here are your prompts and tools, go use them.” They are saying, “Here’s access to a few tools, figure out what you can do with them.”
So I spend a lot of time experimenting, building workflows, and seeing what actually sticks.
When to Use AI for in Project Delivery — and When to Avoid It
What excites me most about AI in project delivery is that we’re finally spending less time being task managers, and more time being actual leaders. That shift is long overdue.
Thanks to AI, we’re finally spending less time being task managers and more time being actual leaders.
I use AI constantly to offload the busywork. It's great at providing answers from a data set of project knowledge, summarizing, drafting notes, and pushing reminders for action items, for example. But you have to double-check it. AI has a bad habit of making up action items or falsifying due dates.
This allows me to focus on guiding the team and solving bigger-picture challenges.
My knowledge of project execution is best applied to strategic planning, anticipating risks, guiding teams through complex scenarios, and driving smarter delivery decisions — not chasing down status updates.
And I won't let AI handle client communication around scope and risk either. There is no replacement for reading personalities, adjusting to communication styles, and drawing on past experience to navigate tough conversations.
Now, if AI could also figure out how to do expense reports and enter time for my team, that would be great. I’m still waiting on that one.
The AI Tools Powering Project Delivery and Workflows
At work, we use a suite of authorized AI tools from Atlassian, Vercel, and Google, plus access to an enterprise tool that taps into different LLMs like ChatGPT and more.
Atlassian offers Rovo, which is great for enterprise searching — it can surface results across Gmail, Docs, Confluence pages, and Jira tickets all at once.
We also use Elvex to create a controlled knowledge base from a defined data set, so the team can query it for project-specific questions.
And Vercel offers V0, which is fantastic for “vibe coding” — it makes it easy to generate and iterate on internal prototypes or visualize ideas quickly.
Outside of work, for my content creation, I rely heavily on ChatGPT, Claude, and a variety of design tools.
Claude has been a game-changer. I no longer need custom plugins or dev support because it will generate front-end code and allow me to preview it, so I can troubleshoot before going live and make my workflow smoother.
And at home, Notion and Notion AI are my brain. They are how I keep my entire family and home life organized.
And I'll say this. AI can't do everything. It hasn't convinced my toddler to eat vegetables. If someone builds that model, sign me up!
The Must-Have AI Tool for Project Managers
I have two favorite AI tools right now.
From a project leadership perspective, it's Elvex. Giving teams the ability to query a database and get answers via chat saves me hours every day.
From a project leadership perspective, it’s Elvex. Giving teams the ability to query a database and get answers via chat saves me hours every day.
You would be shocked at how often a PM is the person people ping when they cannot find or remember something. Now I can just point them to a chatbot. We've created Elvex integrations with Slack so teams can ask the “Project” bot questions about goals, deadlines, team members, or other project details — and get instant, accurate responses.
From a content creation perspective, I have to give credit again to Claude. It has eliminated the need for so many plugins and backend scripts I used to pay for, and it saves me hours by simplifying both front-end and backend tasks.
How AI Is Transforming Core Delivery Rituals
AI is fantastic for reducing information overload at the start of a project.
AI is fantastic for reducing information overload at the start of a project.
Instead of long knowledge-transfer meetings, teams can query a database of project information and get what they need in real time.
It is also increasingly valuable in validating delivery outcomes — from assisting with code reviews to supporting automated testing.
AI does not replace expert validation, but it accelerates the process and helps us catch issues earlier.
Using AI Workflows to Manage Complex, Multi-Stakeholder Projects
On large, complex programs with multiple departments, vendors, and client stakeholders, automated workflows and integrated tools are absolutely critical.
For example, I've set up workflows that send automated reminders in Slack, generate AI-powered daily summaries to reduce meeting time, and sync project plans across different platforms.
Slack recently released its own summary feature for channel conversations, which has been incredibly useful. And when I want to make sure the team is up to date on what’s happened in Jira, I use a Rovo agent to summarize changes to targeted fields on our boards.
Rovo is also becoming more valuable for identifying project trends, which we can then tie into retrospectives and learnings — here's an Atlassian article on how to do that.
This has been especially valuable when clients use different ecosystems, such as Microsoft or Google. Having workflows and AI integrations bridge those gaps keeps everyone aligned without forcing people into a single tool.
This has been especially valuable when clients use different ecosystems, such as Microsoft or Google. Having workflows and AI integrations bridge those gaps keeps everyone aligned without forcing people into a single tool.
Experimenting With Agentic Workflows to Boost Delivery Efficiency
We are experimenting in every area we can — including with agentic workflows. We have developed our own orchestration tool and continue to test different ingestion points.
The focus has been across all phases of delivery:
- Strategic insights
- Design documentation
- Project monitoring, and
- Engineering delivery
We are still iterating, but we expect to see significant efficiency gains as these systems mature.
Why Integrating AI Tools Into Delivery Workflows Remains Challenging
With all that said, it is still way harder than it should be to set up integrations for what I know are common business problems.
AI will happily walk me through the steps, but I would love it to just do the setup for me. Getting it to execute across systems such as Google or Microsoft is another story, thanks to siloed ecosystems.
Another surprise — or maybe frustration — is the sheer variety of models and tools. You really do get wildly different answers depending on which tool you use and its specialty, so staying on top of them and testing them against real use cases takes time.
Why AI Could Create a Future Leadership Gap in Project Management
Project managers are dot-connectors and communication pros. In five years, I think AI will handle 90 percent of the data gathering and analysis we do today.
That means PMs will need to lean even harder into leadership, client management, and people skills — but they aren't going to be out of jobs; the jobs will just be different.
Project managers aren’t going away. The jobs are just going to be different.
The biggest challenge between now and then is the entry-level talent gap we are going to see.
If AI can do all of the busy work that we previously let interns and associate-level PMs tackle, we have to create new associate-level roles for project managers — possibly database and prompt-related — to ensure we have a pipeline of candidates to become future senior leaders. Otherwise, there will be nobody to step up into the leadership roles.
I also think resourcing will look very different. Instead of static roles, we will see short-term needs for things like prompt engineering, AI validation, or integration setup.
Roles will flex and evolve project by project, and PMs will need to be experts at building and managing those mixes.
Advice for Delivery Leaders Adapting to the AI Era
Do not overthink it; just start using the tools. Everyone looks like they know what they are doing, but most people are figuring it out as they go.
Everyone looks like they know what they are doing, but most people are figuring it out as they go.
You are not behind. We all have access to the same AI. The difference is whether you are making time to experiment and find what actually works.
Need expert help selecting the right Other Software?
If you’re struggling to choose the right software, let us help you. Just share your needs in the form below and you’ll get free access to our dedicated software advisors who match and connect you with the best vendors for your needs.
Follow along
Stay connected with Lauren on LinkedIn and her website. And keep an eye on the evolving work her team is doing at Code and Theory.
More expert interviews to come on The Digital Project Manager.
