About Dylan Berry | neurex.dev

The short version

Dylan Berry is the founder of neurex.dev, where he trains teams and leaders to wield agentic AI tools with consistency, reliability, and precision, and helps executives set the AI strategy around them. He builds the curriculum, delivers the workshops, and coaches the executive sessions.

His approach comes from more than 20 years of technology consulting. Most recently he designed and built secure cloud data platforms on Azure and Databricks for enterprise and public-sector organizations in maritime and logistics, financial services, insurance, telecom infrastructure, aerospace, and cybersecurity software. That work spanned landing zones and hybrid networking, infrastructure as code, CI/CD, data engineering and governance, and the security, monitoring, and cost controls that keep platforms healthy in production. Earlier, he led development teams in capital markets and built cross-platform mobile apps.

That background shapes what he teaches. Agents make code, content, and strategy cheap to produce, so the discipline around them matters more: verification, governance, and repeatable process. He builds with coding agents every day, along with the tooling that keeps them honest: verification gates, evaluation sets, and telemetry on cost and context usage.

He has trained professionals and practitioners through formal workshops, internal training programs, hackathons, community events, and one-on-one mentoring. He founded the Toronto Mobile .NET Developers, a community of more than 800 developers, and served as volunteer advisor to Centennial College's Mobile Applications Development program.

Based in Toronto, he writes about what works at dylanberry.com.

How the training is built

Understand the team first

Every engagement starts with discovery sessions that look at your stack, your tools, your constraints, and the work your team actually does. The curriculum is then shaped around that, which is why the syllabus you receive is customized rather than generic.

Practice on real work

Participants bring tasks and features from their own backlog. By the end of the second day, the work they practiced on is work they can ship, and the starter kits travel back with them.

Verification before velocity

The curriculum teaches agentic tools alongside the discipline that keeps their output trustworthy: deterministic tests, verification contracts, review workflows, and deployment gates that catch problems before production does.

The tools your teams already have

Training covers the leading agentic platforms across development, knowledge work, and data engineering, and it teaches principles that transfer when the tools change.