cbzehner
I’m Chris, VP Engineering at Fazeshift. We build AI agents that do accounts-receivable work: billing, collections, payments, and cash application. We're nine engineers plus our Founder/CTO and me, and I'm hiring someone who can help us do more, better: increase both our production capacity and the quality and sophistication of the software we build.
We think AI changes what’s possible here. Instead of giving finance teams another dashboard to stare at, we’re building agents that actually do the work across billing, collections, payments, cash application, and the workflows connecting those systems.
The interesting engineering constraint: we’re intentionally not the system of record. We act across customers’ ERPs, banks, and billing systems. A situation where a timeout doesn’t necessarily mean a write failed, and retrying blindly can be the wrong answer.
A few of the questions we get to work on: * How do you let agents take real actions safely? Permissions, auditability, retries, idempotency, durable execution, and knowing when a human needs to stay in the loop. * How do you build a coherent product when you’re intentionally not the system of record? APIs are inconsistent, webhooks arrive late or twice, and a timeout doesn’t necessarily mean a write failed. * How should an engineering team work when coding agents can take on meaningful chunks of implementation, debugging, testing, and exploration?
We’re especially interested in engineers who have already changed how they work because of AI. If you delegate meaningful work to agents, keep independent tasks moving in parallel, debug where they get stuck, and shave the occasional yak to get things working properly, you’ll feel at home here. We’re still figuring this out too (and updating our priors regularly).
At the same time, generated code is still your code. We care about judgment, debugging, systems thinking, taste, and understanding what you’re shipping.
Our approach is roughly: design enough to find the shape of the problem, then build to learn. There’s a short distance here between noticing a problem, building the fix, and seeing the result.
We’re looking for 5+ years of professional experience, strong TypeScript/full-stack skills, and production ownership. No fintech background needed.
Stack: TypeScript/Next.js, tRPC, Prisma/MySQL/PlanetScale, Inngest, AWS.
I’m Chris, VP Engineering, and I read these applications myself. Email [email protected] with HN in the subject, your LinkedIn/GitHub/résumé, and a few sentences about something you’ve built and how you knew it worked or how AI has changed your approach. No public code needed; please leave out confidential details.
The fundamentals are very different but it’s a case of convergent evolution - and I’m happy to liberally sprinkle clone’s around my rust code
Redivis: Open Company Handbook
Nushell takes the ideas behind Fish even further by incorporating types other than strings and adding more built-in functionality that’s useful for your average command-line user.
I doubt I’ll go back but both are great! Highly recommend Fish and Nushell!
What was the hardest part?
What did you learn that you weren’t expecting?
"Vaporators? Sir, my first job was programming binary loadlifters—very similar to your vaporators in most respects."
―Owen Lars and C-3PO
Zero to Production #3: How to Bootstrap a Rust Web API from Scratch
Show HN: A simple Rust CLI for debugging transient errors
Suggested alternatives to the terminology master-slave:
Primary-secondary, Leader-follower, Active-standby, Primary-replica, Writer-reader, Coordinator-worker, Parent-helper
Suggested alternatives to the terminology whitelist-blacklist:
Blocklist-allowlist, Block-permit
IETF: Terminology, Power and Oppressive Language
A Comic Strip Tour of the Wild World of Pandemic Modeling
When it came time for performance reviews, not only did it make for entertaining reading but it was easy to find all the small improvements they made that normally slip through the cracks.
I solve this problem in most meetings by positioning the main view as a smaller window directly under my webcam. This helps put my focal point close to the speaker from their perspective.
Interview Cake lowers pricing to help Engineering Candidates
Oxidize – Remote-Only Embedded Rust Conference [Happening Now]
Really interesting offer. Do you find new subscribers take you up on this a lot? Seems like a great way to improve your migration tooling!
Integrating E-Signatures into Legal Workflows
Architecting a Roles and Permissions System Using Rails, GraphQL and React
As someone broadly unfamiliar with text rendering it's not clear to me. An introduction to these points would be very helpful