Independent consultant helping engineering teams cut incident noise, close ownership gaps in undocumented systems, and move legacy monoliths to cloud-native architecture — without breaking what's already running.
VPs of Engineering, CTOs, and Heads of Platform at companies where a single backend service is business-critical — and the on-call load, ownership gaps, or migration debt around it have become unsustainable.
This service handles a huge volume of traffic and keeps having OOM-related restarts. Alarm fatigue is burning out the on-call team.
We inherited a complex, undocumented system that only one engineer ever understood — and we need the rest of the team productive on it, fast.
We need to move this legacy monolith to a cloud-native, containerized architecture without a big-bang rewrite or downtime.
I take personal ownership of the system: diagnosing root causes in JVM/GC behavior, observability noise, and architectural bottlenecks, then fixing the highest-leverage issues first.
You work directly with me end to end — from diagnosis through production — and the team is left able to run the system independently, backed by real documentation and structured knowledge transfer.
Owned Pulse, a metric-ingestion service handling ~100,000 req/min from hardware scanners and smart NICs; extended it to add ~50,000 req/min of new input, unlocking a deal-critical integration for a major enterprise customer.
Eliminated recurring OOM-induced service restarts through JVM heap analysis and G1GC tuning — zero configuration-related incidents since rollout — while driving a 65% cut in Sev2 alarm noise.
Ran end-to-end knowledge transfer on an undocumented, business-critical codebase, ramping 10+ engineers to full productivity and turning a single-point-of-failure system into a team-owned platform.
Led a ground-up refactor of a legacy product from 5,000 to 2,500 lines (50% reduction), delivering 60–80% performance improvement with zero functional regression.
Designed and built the end-to-end AWS deployment pipeline (dev → stage → prod) using Terraform, standardizing infrastructure provisioning across environments.
A free 30-minute conversation to understand the problem and whether it's a fit — no pitch, no obligation.
Fixed-fee or day-rate, typically 4–12 weeks, with a clearly defined deliverable agreed upfront.
Architecture review, fix plan, or staged migration delivered with documentation — your team owns it going forward.
Pricing is scoped per engagement based on complexity and duration. Day-rate engagements are available for teams that need focused, time-boxed diagnosis or delivery rather than an open-ended retainer.