Ekho · New York
Software EngineeringR&DFullTimeposted 3w ago
Full-Stack Engineer
New York City HQ ᐧ Full time ᐧ On-site ᐧ R&D ᐧ $160K-$200K + meaningful equity
Buying a vehicle online is pretty sweet. If you disagree, then you probably aren’t a good fit for Ekho; please return to your rock, luddite. For those remaining: the numbers support our belief. Carvana will sell a record high of 800,000 cars online in 2026 (making up around 2% of the total number of used cars sold in the US per year). Moreover, studies show that online vehicle shoppers - like Carvana buyers - report the highest satisfaction rate out of all shoppers. Pretty compelling stuff if you ask us. Thus, our conclusion: buying a vehicle online is pretty sweet.
The obvious followup is ‘why don’t more vehicles get bought online’? Well, it turns out it’s pretty hard to do. Just check out your local dealership website - you’ll most likely see ‘Get My Quote’ buttons on inventory pages rather than ‘Buy Now’ buttons. The horror!
To many people’s surprise, dealers don’t forgo ‘Buy Now’ buttons because they hate the idea of online sales. Quite the opposite: dealers are savvy businessmen and women who can see the success of Carvana and want to emulate their tactics. The true reason blocking the ‘buy now’ button is boring (and hard): taxes, financing, insurance, titling, and registration differ by state and county and most dealers have no reliable way to sell across state lines, so they don’t.
We built Ekho to solve this problem for dealers and manufacturers. Using our technology, a buyer can find a vehicle they like on a dealer or manufacturer site, click ‘Buy Now’, do financing and insurance online, and the vehicle will show up at their door a few days later. The dealer barely has to lift a finger.
Getting here took 24 months of untangling county-by-county titling rules, getting licenses in states Rowan (our South African cofounder) can’t pinpoint on a map, and hooking into 50 state DMV systems, some of which still prefer faxes to APIs. Building this compliance foundation was not sexy, but we’re now on the other side. Now, we get to do the fun stuff: we're adding AI, moving into cars (we started with motorcycles), and are making the whole thing work in the store as well as online. The moat here isn't code. It's the compliance framework, the DMV relationships, and the licenses we spent years accruing. Your move, Anthropic.
We're 34 people in a Flatiron office four days a week - nine of us engineers - with a Claude Code budget only Nadim polices. He can’t catch all of us. If this at all sounds interesting, come talk to us.
We’ve got two primary problems to sort through at the moment.
Cars - we’re aggressively growing in this vertical and we’ve got our work cut out for us. We’ve mastered powersports (motorcycles, dirt bikes, side-by-sides, atvs, etc) and are graduating to the big brother, automotive. We’ve signed some big-wig auto clients who are onboarding shortly and will introduce a step-function change in automotive volume on the Ekho platform over the next 6 months. Top of mind projects to support this growth in volume include revamping our lender architecture, expanding our AI-driven titling and registration protocols, fleshing out our notification service to be multi-modal and more configurable, greatly expanding our trade-in functionality, productionizing lease support, and much more. We’ve got a 12 month roadmap defined. We must execute in 6.
AI Sales Agent - what started as a company hackathon project has turned into the wedge product we use to get our foot in the door of a new client. The sales agent lives as a widget on dealer or manufacturer sites and can manage entire sales workflows from inventory discovery through financing pre-approvals and service appointment bookings. The widget needs to inject itself seamlessly into any third party or custom site, needs to be optimized for high quality lead capture, and eval suites must catch high stakes discrepancies between facts in our data and what we share with buyers. The tool is backed by highly complex inventory data ingestion and management pipelines and, unlike competitive products, buyers can actually buy their vehicle.
Rowan grew up in South Africa, where his dad owned a used car dealership. Chris grew up in Atlanta around some of the biggest dealer operators in the Southeast. They met at Stanford, spent time at Duolingo and Meta respectively, then did YC with the intention to digitize the automotive industry.
The founders are great (and both technical) - but our strength comes from the collective crackedness of everyone else. Bongi, our founding engineer, has known Rowan since high school and turned down a few of his earlier ideas before finally saying yes to Ekho. This guy’s a machine - no other way to put it. Nadim runs AI agents on a server farm consisting of every laptop he's ever owned. Best platform/devops engineer you’ll meet. Rodrigo is the definition of a product engineer. He has the taste of an exceptional PM and his technical depth is highlighted by the persistence of the state machine he designed years ago that we still use to manage the orders flowing through our system. David has architected the cash, ACH, collateralized lending, crypto, and BNPL payment engine that underpins every product. Tommy built the company “brain” at DoorDash and knows exactly how Anthropic’s latest model harness can be manipulated to give us whatever we need. Connie has wrestled our messy, complicated platform into a beautiful product experience through her masterful designs. I could go on, but there are word count limits on these postings. Know there are a handful of other, equally impressive engineers (and non-engineers) who you’ll get to work with and learn from should you join Ekho.
Someone strong across the stack who can be in React one day and in the backend the next, and who takes a rough problem and turns it into something shipped without waiting for a spec. You should care how the thing feels to use, not just whether it passes tests, and you should already be building with AI rather than meaning to start.
The work has real constraints: regulated markets, legacy systems, 50 sets of rules. If that sounds interesting rather than annoying, you'll like it here. The team tends to start around 8:30 and dinner shows up around 7 for whoever's still in.
Stack: React, Node.js (serverless), Express.js, NoSQL
Tools: GCP, Firebase, Stripe, and various SaaS platforms
$160K-$200K base
Meaningful equity
Health, dental, & vision
401(k)
Free lunch and dinners
$700/year work setup stipend
Annual team offsite
First, a call with Bongi to learn about Ekho and walk through something you've built, then a live coding assessment with Chris. The onsite has two technical challenges: build something from scratch with AI tools on, and debug an existing system with AI off. After that there's a systems architecture conversation, lunch with the team, and a chat with Rowan. We move quickly, and we'll be honest with you if it isn't a fit.