Twelve years of scaling, securing, and operating systems under real load at AWS, Amazon, and Unwrap.ai. I give you an honest read on what you've built — where it breaks at 10x, how it fails, and how to catch those failures before your customers do. When the easy fixes are gone, it usually isn't a rewrite; it's the three decisions that are actually costing you.
At AWS I spent three years on the Security Token Service, which is the credential layer every other AWS service authenticates against. When your service sits in that path, you learn what operating at scale actually costs: what has to be measured, what breaks at the tail rather than the average, and how much of reliability is decided long before the incident starts.
At startups I've had the opposite constraint. You have weeks, not quarters, and the architecture has to survive a business that keeps changing underneath it. I've been the engineer re-laying the track while the product kept shipping.
Most of my work lives between those two. I'm painfully familiar with how systems fail, how to catch failures before your customers do, and what to do about them at 2am — and I've learned which of the expensive, careful things are worth doing at your size and which are not.
Most recently I was a Principal Engineer and manager at Unwrap.ai, leading a team of 5 in building Plaid for 3rd party data APIs.
Co-Founder · Unwrap.ai“Jordan was one of the earliest engineers at Unwrap. He led critical transformations of our infrastructure that took us from an early stage startup to an enterprise-grade product that could power feedback intelligence for the largest brands in the world… complex and high-stakes technical decisions that Jordan got right.”
Senior Software Engineer · Alexa“Jordan and I were teammates for multiple years while working together within the Alexa group at Amazon. Jordan brought a strong systems design background to the team that helped guide us through multiple major feature rollouts, which translated to directly improving Alexa's answering capabilities for customers. He set the team's standard for infrastructure programming, and mentored others on how to maintain a high bar of operational excellence for the services we developed. He always had a strong sense for prioritization, and how to make tactical tradeoffs against deadlines...”
Staff Software Engineer · AdQuick“Sharp engineer, low ego, and easy to work with even when things got messy. He asked good questions, took feedback without getting defensive, and shipped clean well tested code. I'd happily work with him again.”
Senior Software Engineer · Amazon“Jordan is a great leader, mentor, and honestly one of the best engineers I have ever worked with. When we were at Amazon, he took charge of a big, complicated project and brought the clarity our team really needed. Thanks to his background with AWS, he helped us set up solid, scalable best practices. He personally helped me understand how things work when a system actually needs to scale, and he always did it with kindness and professionalism. Jordan has a great balance of technical skill and a genuine drive to help people. He was a true inspiration to me both inside and outside of work, and I highly recommend him.”
Engineering Director · Bugcrowd“I had the opportunity to work with Jordan during my time at a OneLogin, and he consistently stood out as a strong back-end engineer. He was very competent technically, with the kind of practical problem-solving mindset that makes a real difference when systems need to work reliably and scale smoothly. What I remember most is his ability to deliver results even under pressure. Jordan often put in extra time—sometimes working late into the night—to push projects forward, resolve blockers, and ensure the work shipped with quality. He wasn’t just focused on getting tasks done; he took ownership of outcomes, and his engineering approach reflected that...”
Enterprise customers were importing millions of rows that had to be analyzed in seconds. The existing infrastructure would stall for thirty seconds and then time out, leaving them looking at a blank page. I replaced batch reporting with a streaming pipeline on Debezium CDC, Kafka, and ClickHouse that ingests 100,000 rows per second and analyzes millions in seconds — taking p90 query latency from fifteen seconds to two and turning a laggy product into a real-time one. Several seven-figure contracts followed.
Unwrap.ai · Real-time analytics platformEnterprise buyers wouldn't approve the platform without proper permission controls for members in their account. I designed and built role-based access control across the customer-facing APIs, which cleared the objection and unlocked six-figure contracts.
Unwrap.ai · Multi-tenant RBACI was the sole engineer on the Security Token Service build for us-gov-west-1
(sts.us-gov-west-1.amazonaws.com). STS issues the credentials every
other AWS service authenticates against, so its arrival is what let other AWS teams
deploy into the region — and ultimately what let the region launch.
Alexa's question-answering degraded every time its underlying data was redeployed. This was a routine internal operation that manifested to users as the product failing to answer their questions. I designed and implemented the deployment algorithm that eliminated it, holding availability through every push.
Amazon · Alexa QnAA senior outside read on what you've built: monitoring gaps, cost, where it breaks at 10x, what's load-bearing that shouldn't be, what to fix first, and what to leave alone. I provide a written assessment with architecture recommendations and projected costs.
Throughput, latency, and cost at the point where the easy fixes are gone. I've built real-time pipelines on Kafka, Debezium CDC, and ClickHouse that took p90 query latency from fifteen seconds to two, and re-architected write paths to drop user-perceived latency by 90%. Usually it isn't a rewrite; it's finding the three decisions that are actually costing you, and knowing which of them you can leave alone.
Most outages are not surprises so much as things nobody was watching. I map how your system actually fails — not the failures that are easy to instrument — and build the alerting, dashboards, and runbooks that catch those modes early enough to matter. This includes the unglamorous parts: what pages a human at 2am, what shouldn't, and what the on-call engineer is supposed to do when it does.
Authentication, authorization, and multi-tenant isolation — so one customer's data can never surface in another customer's account. Encryption in transit and at rest, data protection policy, and the controls an enterprise security review actually asks about.
I spent three years on AWS's Security Token Service, built the OAuth API and multi-factor authentication at OneLogin, reviewed application security and data privacy compliance at Amazon, and designed the role-based access control that let Unwrap.ai sell upmarket.
You describe the problem. I tell you whether I'm the right person. Free, and I'll say no if the answer is no.
Fixed scope, fixed fee, usually less than one week. You get a written overview of your current architecture, a diagnosis of the most critical issues, and a prioritized plan.
I build alongside your team, not around them. Fixed scope, billed hourly.
A couple of sentences is plenty — what you've built, and what's starting to hurt. I read every one of these myself and usually reply within a day.