Atlanta, GA
Jef Aldrich
Technology Leader | AI Transformation | Cybersecurity Strategist
Turning Emerging Technology into Enterprise Outcomes
Atlanta, GA
Technology Leader | AI Transformation | Cybersecurity Strategist
Turning Emerging Technology into Enterprise Outcomes
About Me
With over 15 years working in the financial services industry, I lead global IT teams spanning the US, Europe, and India. My work sits at the intersection of enterprise leadership and hands-on technical development, from architecting Zero Trust security transformations and cloud identity modernization to building AI agents and automation tools in Python.
I've led 20+ M&A integrations, managed IT budgets, and earned certifications including CISSP, CISM, and OSCP. In 2026 I completed a Post Graduate Program in AI & Machine Learning at UT Austin McCombs School of Business, where I built real tools using Python, LangChain, RAG pipelines, LLMs, and computer vision. I believe the best IT leaders don't just oversee technology; they build it.
How I Lead
My job as a leader is to set direction, remove obstacles, and build teams that outperform, then get out of their way.
I don't micromanage. I set clear direction, then spend my time removing the obstacles that stand between my teams and results. People do their best work when their leader clears the path instead of standing in it.
I stay deeply technical by design, but I direct that expertise at the complex, high-stakes programs with the most moving parts, where executive-level technical judgment changes the outcome. Everything else, I delegate and trust.
I invest heavily in relationships beyond my own function. Innovation and collaboration don't happen inside org-chart boxes. They happen when technology leaders are trusted partners to the business, not a service desk.
I find each person's strengths and leverage them, invest in learning and development, push people beyond their comfort zones with stretch assignments, and pair rising talent with stronger teammates. Building the next generation of leaders is the job.
Leading teams across the US, Europe, and India has taught me that global leadership isn't one playbook applied everywhere. It's understanding the cultural nuances of each region and using them to bring out every team's distinct strengths.
I'm a strong believer in continuous education, and I model it. From completing a post-graduate AI & ML program while leading global teams to ongoing leadership development, I hold myself to the same growth standard I set for my people.
By the Numbers
Writing & Insights
Perspectives on leadership, technology, and the journey of building at the enterprise edge.
Reflecting on the IT internship that started it all, and the career that grew from it.
Read on LinkedInWhat I learned building real AI tools in Python, LangChain, and LLMs as a practicing IT executive.
Read on LinkedInBuilding with AI
Production AI systems built to eliminate manual work, accelerate resolutions, and put intelligence where it counts.
An autonomous Python agent that works the IT service desk queue alongside human analysts across five teams. It polls ServiceNow assignment groups on a recurring schedule and, for each new ticket, uses an on-premises LLM to determine what the requester actually needs, then matches it against a library of 30+ skill files that define what the agent can answer and how. The response posts directly to the ticket as a work note within minutes. When nothing matches, it falls back to RAG search across the knowledge base. Skills go further than text: they execute PowerShell and Python automations such as access checks and enrollment workflows, or route the ticket to the team that owns it.
An org-wide AI agent built into Microsoft Teams on the Microsoft 365 Agents SDK, backed by an enterprise LLM gateway running Claude on AWS Bedrock. Employees chat with it naturally, and hashtags route a request to a specialized skill: #kb answers IT questions from the ServiceNow knowledge base using retrieval-augmented generation (RAG) with cited sources, #snow looks up live tickets, incidents, changes and servers in ServiceNow, and a built-in directory tool answers "who is..." questions from Entra ID. Its skill set is extensible without code: each hashtag skill can be an n8n workflow registered in a lookup table, so adding a capability means building the workflow and adding one row, with no source changes and no redeploy. Workflows can answer instantly or reply later through the agent's notification API, which also pushes Adaptive Card approvals, reminders and forms to users and captures their responses for downstream automation.
Monitors the software request ticket queue, extracts the package name, and queries NIST NVD, CISA, and other vulnerability databases for recent CVEs. Findings are passed to an LLM with internal approval rules to generate an allow/deny recommendation. The full assessment report is written back to the ticket for human final review.
Credentials
Validated expertise across security, leadership, and applied machine learning.
Certified Information Systems Security Professional
Certified Information Security Manager
Offensive Security Certified Professional
Kennesaw State University · 2008
UT Austin McCombs School of Business · 2026
The Ken Blanchard Companies · 2026
Microsoft · View Credential ↗
Microsoft · View Credential ↗
Microsoft · View Credential ↗
Projects
A selection of tools I've engineered, from enterprise security GUIs to IoT infrastructure.
A PowerShell-based GUI for launching CyberArk applications, streamlining privileged access workflows for enterprise environments.
View on GitHubA C# implementation of the CyberArk Application Launcher GUI, built for performance and enterprise deployment.
View on GitHubIoT motion detection using Arduino MKR WiFi 1010 with a PIR sensor publishing events over MQTT, an early exploration of connected infrastructure.
View on GitHubGet in Touch
Whether you're exploring collaboration, looking for a strategic IT perspective, or just want to talk about AI and cybersecurity, I'd love to hear from you.