Leaving Ramp

November 22, 2024

Reflection on Day 178.

Today is Day 178 for me at Ramp, and it is also my last day at Ramp. Originally, I started writing this post to describe what I've learned and how I grew as an engineer, but honestly, I’d rather just recap the time I've been able to spend at Ramp.

I joined Ramp as an intern on the Applied AI team, a team that was initialized from the Ramp <> Cohere acquisition. Up until that point, I had never worked at a company with more than 50 full-time employees, so I didn’t know what the culture or work would be like.

The Applied AI team made a big company like Ramp feel like a startup, or a group of friends. I joined when the team was just 8, and half of them were from Cohere. No one was blocked by formalities or procedures, they all just built & shipped their work daily.

Despite the size of Ramp’s engineering team and monolith, I was surprised to see that engineers shipped to production daily. Prod deployments were made multiple times an hour, and engineers were trusted to deploy these changes themselves (with peer reviews). I was able to onboard and ship a small feature to production on my first day.

What I loved a lot about Ramp was the velocity of project execution on engineering. In my first two weeks, I shipped multiple improvements to our fast-retrieval DB security and embedding systems, an internal AI podcast proof-of-concept, and started a new cardholder-facing feature for LLM-automated accounting coding. These projects were barely scoped, if scoped at all, and engineers took lead on both the timeline and execution of work.

As I “ramped” up, I was given as much ownership as I wanted. I asked my team for a model training project, and that led me building an improved version of our fine-tuned transaction embeddings (shoutout Anton!). I asked my manager for more technically-challenging projects, and that led to me working on platform optimization projects, where we shipped 10-20x memory and latency improvements on high-volume codepaths like our transaction search and reimbursement endpoints. In my last month of my internship, I was put in charge of leading our Google Calendar integration and shipping all of its downstream features.

After the summer, I transitioned into a full-time role and continued to maintain the projects I owned, as well as ship new ones. I worked on our AI accounting coding feature from inception to our general access release, continued building and marketing our Google Calendar integration, and continued contributing to our platform optimization efforts. I shipped new projects for our merchant data normalization efforts, better internal data platform tooling, and helped transition more of our codepaths to better infrastructure.

A lot of people ask me what makes Ramp and Ramp engineering so high-caliber, and it really comes down to two things: people and values. Ramp recently hosted “The Art of Hiring” at the NYC HQ with Khosla Ventures, and Eric talked about hiring “Super ICs.” It’s clear that Ramp has hired many super ICs — engineers that can own projects from end-to-end and maintain them are common throughout teams. Ramp is also extremely clear on their values of being customer-centric, growth-oriented, and empowering everyone to ownership.

These two facets combined have built a company filled with incredible ICs and managers, with values that promote work that is high-quality and fast. As I look back on what I’ve gained from Ramp, I've learned a lot about what makes a good company, from the product to the team itself.

Thank you!

My time at Ramp has been filled with nothing but opportunities to learn, build, and ship new things. It wouldn’t have been possible without the people I was lucky enough to work with.

Thank you to my manager, Ter, for pushing me to be a better engineer in almost every way and supporting me in the projects I worked on. I’m grateful to have worked closely with who I think is truly one of the best engineers at Ramp.

Thank you to Yunyu and Rahul, the guys who run Applied AI at Ramp, for taking a chance on me and making my time at Ramp both impactful and super fun.

Thank you to my mentor, Anton, for being a great person to work with on the ML projects I contributed to and an incredible teacher when it comes to building production-level AI features.

Thank you to Karim for introducing me to Ramp and the Applied AI team, and continually supporting my tenure at Ramp.

And finally shoutout to the Applied AI team, Josiah, Cooper, Henry, Roland, and all the friends I made along the way :)

© Made in California.