Brian Taylor
Principal-level software architect and systems engineer. The consistent work across a long career has been turning architectural uncertainty into practical, accountable decisions — and delivering the least complexity that meets the goal.
Resolving Architecture is the practice this work is done under, spanning enterprise integration, modernization, scalability and data processing, security and identity, and decentralized, resilient systems.
25+ years
Principal-level architecture and delivery across startups, growth-stage companies, enterprise systems, regulated environments, and decentralized systems.
Architecture depth
Enterprise, solution, application, integration, security, data, and platform architecture — grounded in hands-on work in Java/Spring, Python, and Rust.
US Navy nuclear
Former naval nuclear reactor operator and prior DoD Secret clearance holder. Honorable discharge. High-consequence operational discipline that still shapes the work.
Selected work
A few representative examples. Each is written around why one direction was chosen over another, not a list of technologies.
Collaborative Bitcoin custody, where reliability, data sensitivity, recoverability, and operational accountability are central. Delivered an initial enterprise architecture charter, a lightweight governance model, role mapping, data classification review, and domain-boundary recommendations — enough structure for custody-grade risk without a heavyweight program.
Banking and retail marketing analytics constrained by Greenplum throughput. Parallelized task processing for a 40% throughput gain on the same hardware, then prototyped a real-time analytics engine sustaining ~12k transactions per second at low cloud cost — opening a new business channel.
Identity and access integration across a large, heterogeneous national healthcare environment. Led design, prototype, and pilot work using SSO, SAML, XACML, RBAC, PKI, and Kerberos — because application-by-application access could not scale across organizational boundaries.
Legal discovery under court-driven time pressure. Designed a distributed processing architecture with remotely deployable discovery agents and central review, achieving more than 10x throughput on the same hardware.
Point of view
Diagrams, documents, standards, and reviews are supporting mechanisms. The central work is clarifying what must be decided, why it matters, who owns the decision, and what trade-offs are accepted — then keeping that reasoning usable while implementation continues.
Research
Meridian — a prototype testing whether structured decision-capture helps teams keep architectural intent intact as AI accelerates delivery.
1M5 — resilient, infrastructure-independent communication that keeps working under denied or degraded network conditions.
SEDA bus — the same minimal staged-message-system design built across four runtimes (Rust, Java, Python, TypeScript) to see what each concurrency model does to it.
All are early or open-source, not commercial. Each exists to find out whether an approach actually helps, not to sell it.
See the researchRésumé & contact
Open to conversations about senior architecture and engineering work, past projects and references, or the research.