Podcast

AI vs. Budget Blowouts: Catching Capital Risks 200 Days Early

Brad Gyngell
Author

AI Meets Capital Project Risk Management

Capital projects are notorious for budget blowouts and delays. What if AI could predict financial distress 200 days in advance and prevent multimillion-dollar cost overruns?

In Episode 6 of the Hevi AI Podcast, Paul Culvenor and Brad Gyngell sit down with Cuong Quang, co-founder of Octant AI, to discuss:

Why 80% of major projects run over budget and how AI can fix it
How machine learning can detect financial distress before it happens
The biggest AI trends impacting mining, construction, and infrastructure
The future of AI-driven project forecasting and risk management

Cuong has over 20 years of experience managing mega-projects across mining, oil & gas, and infrastructure for Rio Tinto, BHP, Chevron, and Total. His company, Octant AI, uses AI to improve cost forecasting accuracy by up to 87%.

This episode is packed with insights for executives, project managers, and finance leaders who want to de-risk major capital projects and turn AI into a competitive advantage.

From AI Research to Industry Transformation

Cuong’s AI journey began 25 years ago, just after IBM’s Deep Blue defeated chess grandmaster Garry Kasparov. He was there when AI was still rule-based expert systems—long before machine learning, deep learning, and today's large language models.

After studying at Oxford, Cuong shifted his focus to one of the biggest problems in project management—cost and time overruns.

“I went to Oxford to figure out if the cost blowouts I’d seen in my career were an industry-wide problem. Turns out, they weren’t just common—they were everywhere.” — Cuong Quang

This led to the founding of Octant AI, a company using machine learning to predict financial distress in capital projects.

🔹 Core problem: 80% of major projects experience significant cost and time overruns.
🔹 Solution: AI-driven forecasting that provides early warnings up to 200 days in advance.

Why Do Capital Projects Fail? (And How AI Can Fix It)

The biggest causes of project overruns:

🚧 Overly optimistic planning – Forecasts often rely on manual estimates and historical trends rather than real-time data.
🚧 Poor risk visibility – Projects have too many moving parts, making it difficult to detect financial distress early.
🚧 Data silos & inefficiencies – Critical project data is fragmented across teams, limiting predictive insights.

🔹 Enter Octant AI: AI-Powered Early Warnings

Octant AI learns patterns from past projects and real-time data to forecast financial risk more accurately.

Predicts financial distress up to 200 days in advance
Improves cost forecasting accuracy by up to 87%
Identifies risk factors like contract structure, weather, and project type

“One of our customers avoided a $3.5M loss thanks to an early warning from Octant AI. They had 200 days to take action before the cost overrun happened.” — Cuong Quang

How Does Octant AI Work?

🔹 Step 1: AI Learns from Past Projects

  • Requires only 22 past projects and 6 key data points to start making predictions.
  • Uses small data methods instead of deep learning (ideal for capital projects with limited historical data).

🔹 Step 2: AI Monitors Real-Time Project Data

  • Tracks factors like contract type, weather, regional risk, and financial structures.
  • Continuously updates predictions as new data comes in.

🔹 Step 3: AI Provides Actionable Early Warnings

  • Alerts teams when cost overruns are likely to occur.
  • Helps leaders adjust forecasts and make data-driven decisions.

“Most projects just roll bad performance forward. Our AI shows where you should be—not just where you’re heading if nothing changes.” — Cuong Quang

The Adoption Challenge: Why Are Some Organizations Slow to Use AI?

Even though AI can prevent catastrophic cost overruns, adoption in mining, construction, and infrastructure is still slow.

🔴 Trust Issues – Project managers with 20+ years of experience often trust gut instinct over AI-driven insights.
🔴 Legacy Thinking – Many projects still rely on outdated forecasting models and spreadsheets.
🔴 Data Privacy Concerns – Organizations worry about sharing sensitive financial data with external AI models.

How Octant AI Overcomes These Barriers

100% Secure, Private Models – Customers own their AI models, and data can be deleted after training.
Interpretable AI – Unlike black-box deep learning models, Octant AI explains why it makes each prediction.
Finance & Risk AlignmentCFOs, risk managers, and executives see clear financial value in AI-driven forecasting.

“Once executives see the financial upside—preventing million-dollar losses—it becomes a much easier conversation.” — Cuong Quang

What’s Next for AI in Mining, Construction, and Infrastructure?

🔹 AI-powered project forecasting will become standard – No more relying on gut feel and spreadsheets.
🔹 Agentic AI will automate workflows – AI will handle more complex project management tasks.
🔹 Physics-informed AI will revolutionize simulations – AI will simulate entire projects before execution, reducing risk.

“We’re only at the beginning of the AI revolution. The companies that move fast will have a massive competitive advantage.” — Cuong Quang

Final Thoughts: AI is Reshaping the Future of Capital Projects

AI isn’t just a buzzword—it’s actively saving companies millions by preventing cost blowouts before they happen.

🚀 Want to future-proof your projects?
📅 Book a demo with Octant AI

🚀 Want to see Hevi in action?

📅 Book a demo at https://calendly.com/paulculvenor/30min

🔊 Listen to the full episode on:
Spotify: https://open.spotify.com/episode/1jaqBRK1d4yj85bakVw2G8?si=N1IBR8lyTKaEIQl4u1XfEA
YouTube: https://youtu.be/nD00QqbB0Cc

Drop us a message and see how we can help you!

A headshot of Brad Gyngell
Brad Gyngell
Co-founder & CEO
a headshot of Paul Culvenor
Paul Culvenor
Co-founder

Get in touch with us

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
© 2022 Hevi Pty Ltd
Terms of use
Privacy policy
© 2022 Hevi Pty Ltd