Today’s reliability work isn’t just about machines anymore; it’s about using AI to make faster, smarter calls. Recently, we sat down with Allen Garcia, a mechanical engineer at Baker Hughes and a PhD candidate at the University of Maryland focusing on AI in engineering.

Allen knows both worlds: the nuts-and-bolts side of design, and now, how machine learning can shake up everyday engineering tasks like RCAs, FMEAs, and mining hidden knowledge in work orders.

Quick Overview (TL;DR)

• Allen Garcia works at Baker Hughes while pursuing a PhD in mechanical engineering with an AI focus.
• He breaks down how large language models help engineers tackle piles of messy text and build knowledge graphs.
• He explains how AI speeds up RCAs and FMEAs, giving engineers more time for real problem-solving.
• Allen highlights how EasyRCA’s AI features reflect this human + AI teamwork approach.

Table of Contents

• Allen’s Journey: From Mechanical Design to AI
• How AI is Reshaping Reliability Work
• Allen’s Real-World Insights on AI Tools for RCA & FMEA
• Final Thoughts: Human Engineers Still Matter
• How EasyRCA Supports AI-Driven Root Cause Analysis

Allen’s Journey: From Mechanical Design to AI

Allen Garcia didn’t plan to get into AI when he started out as a mechanical design engineer at Baker Hughes. His early work was hands-on, solving real-world machinery problems every day. But curiosity and a bit of good timing pulled him deeper.

“I always wanted that next level of depth and rigor,” he says. So he went back for a PhD, and found himself drawn into the fast-changing world of machine learning.

How AI is Reshaping Reliability Work

Allen focuses on large language models (LLMs), smart tools that help engineers process massive piles of text, like maintenance logs or old work orders.

“LLMs help us find what’s buried,” Allen says. “They pull out useful insights so you don’t waste hours digging.”

But the payoff goes further: instead of spending days building an FMEA or an RCA, engineers can use AI to build the first version in minutes, then spend the rest of their time validating and applying it.

Allen’s Real-World Insights on AI Tools for RCA & FMEA

Allen sees practical examples of this every day:

• Dynamic RCAs: Imagine an RCA that updates itself as new sensor data rolls in.
• Smarter FMEAs: AI cuts the grunt work of building failure mode effects analysis, freeing time to check, adjust, and improve.
• Better questions: Engineers still need to steer the tool. “AI is fast thinking, not deep thinking,” Allen says.

He’s tested EasyRCA’s AI and likes how it works alongside people. “It doesn’t hand you a finished answer, it helps you get there faster.”

Final Thoughts: Human Engineers Still Matter

One thing Allen makes clear: AI isn’t replacing good engineers anytime soon.

“These models can’t match real-world intuition yet,” he says. “They help us get through the data faster — but the real value is in what we do with it.”

When AI speeds up the busywork, engineers can finally spend more time solving problems that matter.

How EasyRCA Supports AI-Driven Root Cause Analysis

At EasyRCA, we believe reliability will always be about people working with AI, not people replaced by it. Our AI features help teams build cause trees, suggest hypotheses, and handle the repetitive parts faster, so engineers stay focused on solutions.

Want to see how EasyRCA’s AI can help you solve smarter? Explore EasyRCA’s AI features.

To watch the full podcast episode with Allen, click here: YouTube

Further Reading:

• Customer-Focused Reliability: Lessons from Liebherr Mining’s Todd Heindselman
• Transforming Manufacturing Reliability: Insights from Rehrig Pacific’s John Ramirez