Before he'd even gotten Slack set up, Elvin Rivera was already on a plane to Dallas for his first team offsite. By the time he landed back home, he already felt like he belonged. That first week turned out to be a preview of everything that followed: a company that moves fast, trusts early, and builds the kind of environment where an engineer can grow further than they thought possible.
Pre-Abnormal
Before joining Abnormal, Elvin spent three-plus years in tech consulting, where strict guidelines meant customer data could never touch AI tools. Reasonable policy, but it had a real cost to his development.
"It slowed down my learning considerably," he said. "We were limited to general research, and implementation had to be done manually to develop solutions. It felt very linear and didn't help me grow."
When a project landed outside his existing expertise, the path forward was long: research, manual implementation, slow iteration. The tools that might have accelerated his understanding were simply off the table, and that constraint shaped how he carried himself at work.
"I used to get nervous being put on projects where I had little to no exposure to the technologies involved."
The nervousness was practical, not abstract. Being handed a complex system without a fast way to build understanding meant that if someone asked a question he couldn't answer, it would be obvious pretty quickly.
Day One in a Different World
At Abnormal, the difference was immediate. Having access to Claude Code, Cursor, Gemini, and ChatGPT from day one changed how Elvin could move through unfamiliar territory, but what surprised him more than the access was the culture around it.
"There's no unspoken pressure to hide that you used AI while still needing to sound like you figured it out yourself," he explained. "People openly share what they know, what they don't, and use AI collaboratively to find answers together."
That openness extends beyond individual work. His team is actively building systems to automate the regular, mundane tasks: project updates, turning meeting transcripts into action items, managing client requests. “When a team moves in that direction, it alleviates a huge amount of workload and frees up our resources to focus on what actually matters,” Elvin said.
"I've genuinely learned more in my time here than I did in 3+ years at my previous job."
The first time that really landed was when Elvin was asked to make changes in a massive monorepo with dependencies and directories running in every direction. In consulting, that kind of task would have stopped him cold.
"I had AI help me build a mental map of the codebase and point me toward exactly what needed to change," he said. "I moved through it without hesitating once, and that's when it hit me: with AI, you can hit the ground running on almost anything, even tasks completely outside your expertise."
Owning the Whole Thing
That shift in confidence showed up quickly in the scope of what Elvin was trusted to own. The clearest example was a project where he was handed a rough concept and asked to design an entire platform from scratch: infrastructure, CI/CD automation, AI integrations, all of it, with every architectural decision his to make.
"I wore every hat myself, architect, builder, and implementer, and shipped a complex, end-to-end solution on my own," he said. The platform his team now uses to deploy prototyped apps into production.
"Looking back at it, the thought that hit me was: this is what operating at a principal level actually feels like."
A later project reinforced it. Deploying OpenTelemetry for Claude Cowork meant navigating a stack of services that were almost entirely new to him: OpenSearch, Snowflake, OTEL, EKS deployment design. He used AI to understand how everything connected, build diagrams, and fully own the infrastructure deployment. When it was done, he put together a primer series so anyone who came after him could go from zero to understanding the entire solution.
That instinct, to learn and then make it easier for the next person, runs through how Elvin approaches the work. When a teammate wanted to learn Terraform, he didn't point them to documentation and move on. He built a hands-on lab, used AI to handle the actual coding so the teammate could focus on the underlying concepts, and taught him how to ask AI better questions. "Watching their confidence grow in real time was genuinely rewarding," he said. "It reminded me that the best part of this work isn't just solving technical problems. It's the chance to bring others along."
A Team That Moves Forward
The environment Elvin describes goes beyond the tools available to him. It's about the people he's building alongside.
"On my direct team, there's almost no friction. Everyone is heads-down learning, growing, and pushing in the same direction. No politics, no egos, just a shared sense of mission."
He points to the offsite as proof that this culture was built deliberately. Sitting down to dinner and finding your CISO at the same table, with no formality and no distance, isn't something that happens by accident.
"Seeing that human, non-corporate side of leadership made him feel far more approachable. That one interaction gave me the confidence to collaborate with him directly on several initiatives in the months that followed."
That access shaped how Elvin shows up for his own teammates. "When you've actually shared a meal and laughed with someone, you're more invested in their success."
What Comes Next
Elvin isn't looking back at consulting and wondering what might have been. He's watching Abnormal grow into something bigger than email security and wants to be part of what it builds next.
"To see how the company grows beyond the email security offerings and shows the world we're a truly AI-focused company and can build anything, that's what I'm most excited about," he said.
The engineer who used to hesitate in front of complex architectures is the same one now designing platforms end-to-end, leading deployments, and teaching teammates how to build. Getting there didn't require more years of experience, just the right environment.
