Faces of Uncountable
Meet Data Scientist Avi Kejriwal
Avi describes the joys and challenges in designing his favorite feature
We’re happy to share a series of interviews with members of the Uncountable team. The purpose of these is to shed light on the people behind our platform, including their professional backgrounds, what led them to join our team, and what they work on behind the scenes.
Today’s interview is with Avi Kejriwal, Data Scientist at Uncountable.
Employee profile: Avi Kejriwal — Data Scientist
UNC: Can you introduce yourself? What did you do before joining Uncountable?
Avi: Of course! My name is Avi, and I’m a member of the data science and engineering team here at Uncountable. A lot of my previous experience has been in the hardware space, physically assembling computers and servers, doing a lot of electronics testing. And I think what separates Uncountable from a lot of the other places that I’ve seen in the software space is [an atmosphere of] being surrounded by a lot of smart and driven people.
One of my philosophies is that if you’re the smartest person in the room, then you’re in the wrong room. And at Uncountable, I feel like there’s a lot to learn from everyone in the company. It’s very small [team], but [we are] also very driven.
UNC: What stands out to you about the work Uncountable does?
Avi: I think one thing that stands out is the scope of our work. At Uncountable, we work all over the R&D space, and I get to learn about how all kinds of materials across the world are designed: things like rubbers, adhesives, glass, food, the list goes on. And occasionally, we might get samples from materials that we helped design, which makes me excited about some of the prospects and materials I get to learn about.
UNC: What do you do on a day-to-day basis at Uncountable?
Avi: That’s a funny question because one of the things about being at Uncountable, and a small startup, is that my day to day responsibilities are very broad. I can be doing very different things on a daily basis. There’s a mix of developing new features. There’s a mix of responding to customer asks. And, there’s a mix of creating new customer computations or setting up new, customized machine learning models for various customer projects.
UNC: What’s the most surprising or exciting aspect of the job at Uncountable?
Avi: I’ve been at Uncountable for about three years now, and I think that the most exciting thing about the job is the impact and the scope of it. As we build out these features and systems from the ground up, I start to see kind of how widespread the impact I can make in terms of speeding up R&D and speeding up workflow for scientists around the world. As we make these systems better and better, there’s exponential growth in that it gets easier and easier to deliver the same level of impact.
One project that I’m currently leading is one I’ll call “machine testing uploads.” One of the most challenging things for our customers is the actual process of loading data into our software. Nuances aside, oftentimes the challenge is just that there is a lot of data. So, with this project, what we’re doing is constructing a pipeline for users to automatically load and analyze data from their testing equipment. With this, we want to be parsing out key details, and putting them in the right places in our software with a single click. The end result is that analysis and work that normally take five minutes per file in Excel can now be done with a single click.
UNC: What feature are you most proud of working on at Uncountable?
Avi: I think the feature that I’m most proud of would be the pipeline, the pipeline feature, Because it started out as being very customized one off scripts that was like on a very specific test basis. But over time, I’ve seen it evolve into a much more robust, much more flexible system. And pipelines that took hours to build two years ago, now can be done in less than 10 minutes.
UNC: Can you say more about how this Pipeline feature works?
Avi: When a lot of scientists are running tests in the lab, they tend to use very specialized testing equipment. And those pieces of equipment tend to have exports into text or Excel, which includes a lot of raw data that the machine has been collecting over time. Oftentimes, when scientists are analyzing this data, they’re doing it manually in Excel. And they’re doing very specific computations, very specific calculations on it. And they have to do this for each file individually, so it can take a lot of time to even analyze a single test.
So, what we do with this pipeline is automate a lot of that process. All the user has to do is just drag and drop this export file into our system, and it will automatically pull up the calculations they want. It will automatically pull out the logs over time and it will put it in and package it in a way that they can access and search it very easily.
This transcript has been condensed and lightly edited for clarity.
