This AI Startup CEO Is Changing Project Delivery Into a Proactive Discipline With Predictive Intelligence Models
Lloyd Skinner is the CEO of greyfly.ai, where he uses AI to build predictive intelligence that gives project leaders actionable insight before they're hit with unexpected problems and costs. In other words, he's transforming project delivery from a reactive discipline into a proactive one.
We sat down with him to understand how he's doing it — and what it really means in practice. Here's what he had to say.
Moving From Traditional Product Delivery to Ai-Powered Delivery
I'm the CEO of greyfly.ai, where we help organizations deliver projects more predictably and profitably through Artificial Intelligence.
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
In the past, I oversaw traditional project delivery. Having worked across major programs and transformation initiatives, I’ve seen how traditional approaches fall short in managing complexity.
That's why my role has transitioned to enabling organizations to rethink how delivery happens in an AI-first world. I focus on enabling senior project and portfolio leaders to use their own data more intelligently, thus improving outcomes, reducing risk, and building the case for scalable AI in PM adoption.
At greyfly.ai, we bridge the gap left by traditional approaches by embedding AI into project delivery, giving leaders actionable insight before issues arise, and helping them achieve measurable, sustainable results.
How AI Is Turning Project Delivery Into a Proactive, Insight-Driven Discipline
AI is transforming project delivery from a reactive discipline into a proactive, insight-driven one.
AI is transforming project delivery from a reactive discipline into a proactive, insight-driven one.
Rather than focusing on reporting what’s already happened, my attention is now on helping leaders anticipate what’s about to happen: using predictive intelligence to guide real-time decisions and prevent costly overruns.
We’re now using models that forecast project success, identify early risks, and highlight where intervention will have the greatest impact. And that’s fundamentally changing the way project leaders plan, prioritize, and manage resources.
As a result, I spend far less time on manual data analysis, dashboard reviews, or post-event lessons learned. Instead, I focus more on strategic adoption, data quality improvement, and embedding AI capability into project organizations.
My role is increasingly about helping executives move from curiosity about AI to tangible, scalable change, turning insight into measurable outcomes.
Why Data-Led, Lightweight Delivery Systems Outperform Traditional Project Management
We've made a deliberate move away from traditional, document-heavy project-management methods toward lightweight, data-led delivery systems that prioritize agility and foresight over administration.
In the past, like many organizations, we relied heavily on structured templates, manual progress updates, and formal review cycles. That approach created delays, duplication, and often masked emerging risks until it was too late. Over the past year, we’ve transitioned to a more integrated, continuous insight model built around our own Intelligent Project Prediction (IPP) platform and our principle of Minimum Viable Governance (MVG), ensuring control and assurance remain intact, but with far less overhead.
Practically, that meant connecting our existing PPM tools into a single data pipeline hosted in Azure. IPP then analyzes this live data to predict outcomes, flag risks, and highlight delivery trends in real time. This replaced weekly report compilations with automated dashboards that refresh daily and meet agreed thresholds automatically.
We also introduced AI-driven compliance checking, which scans project documentation against governance standards and validates alignment to MVG principles. The discovery and transition took around six months, including data mapping, integration setup, and change workshops with PMO teams to build confidence in both the predictive insights and the governance model.
The results have been tangible:
- Reporting effort cut by more than 50%
- Risk visibility improved dramatically
- Decision-making cycles shortened from weeks to days.
- And, perhaps most importantly, the team culture changed with project leaders now spending less time chasing status and more time solving real issues.
In short, the shift to lightweight, AI-enabled delivery isn’t about abandoning rigor, it’s about redefining governance so that control, compliance, and insight happen continuously; not retrospectively.
The shift to lightweight, AI-enabled delivery isn’t about abandoning rigor, it’s about redefining governance so that control, compliance, and insight happen continuously; not retrospectively.
A Real-World Example of AI Integration Improving Project Delivery Outcomes
A great example comes from our work with a major UK telecommunications company, where we supported a multi-phase AI-in-project-management engagement.
The client wanted to understand how AI could improve predictability and performance across their large project portfolio, but like many enterprises, they faced fragmented data, inconsistent reporting, and a reliance on manual judgment.
We began with an AI Adoption Strategy, then moved into a Discovery Phase. This involved a structured PMO assessment, technical architecture review, and data quality analysis using tools such as Microsoft's Power BI for exploratory diagnostics. Our data science team built predictive models using the client’s own historical project data to forecast likely outcomes, including risk exposure, schedule slippage, and delivery confidence.
The setup took several weeks, including securing data access, defining the schema, cleaning the datasets, and validating the model outputs with project professionals. Once the models were trained, we visualized insights via Power BI through interactive dashboards that highlighted high-risk projects in real time and explained why certain projects were trending off course.
The result was a step change in decision making: Senior leaders could focus their interventions where they mattered most, rather than reactively chasing issues after the fact. This early pilot demonstrated measurable potential to reduce project overruns and provided a clear business case for scaling AI across the wider organization.
It wasn’t all smooth, though. The hardest part wasn’t the technology, but the data readiness and the change management around trusting predictive insights over instinct. But that’s precisely where AI in delivery creates value. It gives leaders the confidence to act on evidence; not hindsight.
How Agentic Workflows in Data Preparation and Governance Validation Accelerate Delivery
We are also actively experimenting with agentic workflows and intelligent orchestration as part of our broader mission to make project delivery faster, leaner, and more adaptive.
Our focus has been on automating the most time-consuming coordination and assurance tasks — the "invisible admin" that slows delivery without adding real value. Specifically, we’re using agentic orchestration to streamline:
- Data preparation: AI agents now source, cleanse, and align project data across multiple systems before predictive modeling begins.
- Governance validation: AI agents automatically test project artifacts against MVG standards, ensuring compliance without manual intervention.
We’ve embedded this orchestration layer into our IPP ecosystem so that these agents operate seamlessly behind the scenes, triggering updates, generating insights, and prompting actions without requiring user intervention.
So far, the results are very promising. We’ve seen reporting-cycle times and data-preparation times reduce significantly. But perhaps more importantly, teams are spending their time interpreting insight rather than chasing information.
It’s still evolving, and we are learning how best to balance autonomy and oversight. But the direction is clear: Intelligent orchestration will become the backbone of modern delivery. It’s not about replacing people; it’s about ensuring every human effort is directed where it delivers the most value.
How Predictive Intelligence Is Reshaping Core Delivery Rituals
Additionally, we are rethinking delivery rituals from the ground up, treating AI as a genuine contributor to decision making; not just an analytical assistant. That means reshaping how we define scope, align teams, validate progress, and execute delivery to make the most of predictive intelligence.
We are rethinking delivery rituals from the ground up, treating AI as a genuine contributor to decision making; not just an analytical assistant. That means reshaping how we define scope, align teams, validate progress, and execute delivery to make the most of predictive intelligence.
- Defining scope: When we define scope, AI helps us quantify complexity from the outset. By analyzing historical data, our models identify similar past projects and predict potential delivery challenges. This gives sponsors and PMOs a quantified starting point so that scope isn’t just defined by ambition, but by evidence.
- Aligning teams: For team alignment, AI now supports the early identification of capability gaps or overextension risks based on workload and historic performance. This allows us to configure delivery teams with far more precision and confidence, rather than relying on static resource plans.
- Validating progress: Validation and assurance have evolved a lot. We use AI-driven compliance checks to ensure project outputs align with governance standards, embedding MVG into the workflow. Instead of waiting for end-of-phase reviews, assurance is now continuous. It’s proactive, lightweight, and data-backed.
- Executing delivery: In execution, we’ve supplemented retrospective status meetings with predictive dashboards from our IPP platform. These dashboards surface early warning signs, highlighting emerging risks, and prompting focused discussions on what matters most. AI doesn’t replace leadership judgment, but it reframes conversations, moving them from reporting history to managing the future.
The biggest change is cultural: AI has become a participant in the delivery process. It doesn’t just automate work. It amplifies intelligence, helping humans make faster, better, and more confident decisions.
Automate Low-Value Project Tasks With AI for Better Efficiency
Right now, the areas most ready for automation and AI support are those that consume the most time but add the least strategic value, things like project forecasting, status reporting, and risk identification.
Right now, the areas most ready for automation and AI support are those that consume the most time but add the least strategic value, things like project forecasting, status reporting, and risk identification.
These processes are rule-based, data-rich, and repetitive, making them ideal candidates for intelligent automation.
However, the human touch remains essential in areas that require judgment, empathy, and negotiation, such as stakeholder alignment, culture change, and strategic prioritization. AI can inform the conversation, but leaders still need to interpret context, manage politics, and make value-based decisions.
In essence, AI should take over the cognitive heavy lifting so that project leaders can focus their energy on influencing outcomes; not managing information.
What It Really Takes to Make AI Adoption Stick in Project Delivery
The big question is: ‘What does it actually take to make AI stick in project delivery?’
The answer is not more technology. It isn't complex algorithms or huge investments. It is belief and behavior — AI adoption is about a mindset shift.
We have learned that the tipping point doesn’t come when the model gets smarter; it comes when people start to act on the insight it provides. That’s why I often say AI adoption is a human change program powered by data, not just a digital one. Once teams see that predictive insight helps them avoid surprises rather than explain them, the shift is permanent.
The real transformation happens when delivery leaders stop treating AI as an experiment and start treating it as part of how they lead. That means creating trust in data, being transparent about what AI is showing, and building confidence in teams to use it in their day-to-day decision making.
And what really stands out is how quickly trust builds once people see the accuracy of the predictions that AI provides. Teams move from debating what went wrong to discussing what’s likely to happen next — and that shift fundamentally changes the culture of delivery.
Build an Intelligent AI Ecosystem for Project Delivery
Our delivery tech stack has evolved significantly over the past year as we’ve deepened the integration between AI models, analytics platforms, and traditional project management tools. At the core sits our own IPP platform, it’s the intelligence layer that analyses project data, forecasts outcomes, and identifies early risks across portfolios.
We utilize Microsoft Power BI for data exploration and visualization, as it remains a practical way to bridge project data from multiple sources while ensuring stakeholders can interact with insights directly. On the data side, we’ve strengthened our use of Azure environments to securely host and scale AI workloads, improving performance and governance.
In parallel, we’ve simplified rather than expanded our toolset. Traditional planning tools such as Planview and Primavera still have their place, but they now serve as data contributors to the predictive models rather than standalone management systems. The emphasis is on interoperability, connecting existing tools so that insight flows seamlessly instead of being trapped in silos.
And we’ve replaced manual, spreadsheet-based risk tracking with automated model outputs and introduced AI-driven compliance checks to validate process adherence. These changes have reduced manual reporting effort by over half and shifted focus from collecting data to interpreting it.
In short, the evolution of our tech stack isn’t about chasing new tools; it’s about creating an intelligent ecosystem, one where existing platforms are enhanced by AI to deliver foresight, accuracy, and measurable value to every project decision.
The evolution of our tech stack isn’t about chasing new tools; it’s about creating an intelligent ecosystem, one where existing platforms are enhanced by AI to deliver foresight, accuracy, and measurable value to every project decision.
Why Project Management Is Evolving From Administrative Work to Intelligence-Led Strategy
Over the next five years, AI will completely redefine what it means to manage a project. The shift won’t necessarily be about adding more tools. It’ll be about rebalancing the relationship between people, data, and decisions.
I believe we are moving toward a world where project management becomes an intelligence-led discipline, not an administrative one. AI will handle the mechanics, such as forecasting, reporting, assurance, while human leaders focus on purpose, priorities, and outcomes. Teams will rely on continuous prediction rather than retrospective control, and governance will evolve toward intelligent systems like our concept of MVP, ensuring assurance happens invisibly, as part of the flow of work.
Within five years, most project organizations will have AI co-pilots embedded at every level, which surface insights, orchestrate workflows, and challenge assumptions in real time. The best leaders will be those who know how to partner with AI, not compete with it.
Ultimately, project delivery will shift from managing uncertainty to mastering foresight. Those who embrace AI early will not only deliver more predictably, but also fundamentally change how value, impact, and success are defined.
Project delivery will shift from managing uncertainty to mastering foresight. Those who embrace AI early will not only deliver more predictably, but will also fundamentally change how value, impact, and success are defined.
Why Data Readiness and Human Awareness Are Critical for AI Success in Delivery
My advice to delivery leaders is to focus on two things above all else: data readiness and human awareness.
Data readiness is the foundation of every successful AI initiative. Most organizations underestimate how much untapped value already sits in their existing project data, but it’s often fragmented, inconsistent, or locked away in systems that don’t talk to each other. Getting that data into shape doesn’t require a massive transformation program; it starts with understanding what you have, improving quality where it matters, and building confidence in its use. Once your data is trusted, predictive insight follows naturally.
The second element is human awareness, helping teams understand what AI can do for them rather than what it might take away. The most successful transitions we’ve seen happen when delivery professionals feel empowered by AI, not threatened by it. It’s about developing new instincts, learning to question patterns, interpret forecasts, and use AI as a strategic partner rather than a reporting tool.
If leaders invest equally in those two dimensions, data readiness and human awareness, they will create a culture where AI genuinely improves delivery performance, rather than simply adding another layer of technology.
That’s where the transformation really begins.
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
You can follow Lloyd on LinkedIn as he transforms project delivery into a proactive discipline. And check out greyfly.ai, where he's making it happen.
More expert interviews to come on The Digital Project Manager.