Tyler Weldon / Founder
Tyler is a cloud infrastructure builder and offensive security engineer whose work spans bare metal, Kubernetes, accelerated compute, and the attack paths between them. He founded Dark Forest to bring that operator depth to continuous offense and defense.
Builder / Breaker
Before testing cloud and AI infrastructure, Tyler built it: load balancers, managed Kubernetes, Slurm-on-Kubernetes, Terraform tooling, storage systems, and production fleets.
That foundation shaped a practical security discipline—trace the real attack path, demonstrate consequence, and stay with engineering teams through remediation. Dark Forest extends that discipline with governed agentic workflows that can operate at the pace of modern infrastructure.
Field experience
Nine years across the stack—from physical cloud infrastructure to multi-tenant AI platforms and the security boundaries that protect them.
Linux Systems Administrator
Grew from operating an international compute fleet, Ceph storage, and physical hardware to leading the design and implementation of Vultr Load Balancers and Vultr Kubernetes Engine.
Infrastructure Engineer
Served as an early security voice before joining as a founding member of the red team. Identified critical platform attack paths—including container escapes, cross-tenant compromise, lateral movement, and escalation to cluster administration—and drove remediation with engineering teams.
Bare Metal
Strengthened supply-chain, firmware, identity, and platform security for bare-metal AI infrastructure. Conducted the company’s first internal penetration test and established external attack-surface monitoring.
Founder
Building focused security and infrastructure capabilities for organizations operating high-stakes cloud, AI, HPC, and sensitive-data systems.
Why Dark Forest
Dark Forest sits at the intersection of infrastructure operations, adversarial thinking, and automation. Tyler has worked at that intersection from both sides.
Deep fluency in Linux, bare metal, distributed storage, networking, Kubernetes, Slurm, and GPU infrastructure means less discovery overhead and more meaningful testing.
Tyler’s red-team work chained weaknesses into real platform outcomes. Swordholder carries that focus into continuous attack-path discovery and exploit validation.
Security is not finished when a report lands. Experience designing platforms and driving fixes keeps recommendations grounded in operational reality.
Years of infrastructure automation inform a deliberate model for agentic security: explicit scope, auditable actions, approval gates, and humans at consequential decisions.
Continued study
The technical landscape does not stand still. Neither should the person assessing it.
GitHub / oogyArizona State University
2026—2027 / In progress
Coursework at Ocean County College
Python, Bash, Go, C/C++, Rust, Nix, Terraform, Ansible, Kubernetes, and Linux
Work directly with the founder