Main Responsibilities of a Product Engineer at Graphistry
Alex, a senior product engineer at Graphistry, works closely with a small team, collaborating on a new LLM application called Louie; Alex's responsibilities span the entire product lifecycle, from "Figma Designs" and prototyping to coding in Python and TypeScript and onboarding new team members, showcasing a blend of technical expertise and leadership within a fast-paced startup environment.
Software Development, Full-Stack Engineering, AI/LLM Application Development, Data Engineering, Product Design
Advizer Information
Name
Job Title
Company
Undergrad
Grad Programs
Majors
Industries
Job Functions
Traits
Alex Warren
Product Engineer
Graphistry
University of Arizona 2015
None
Computer Science
Technology
Product / Service / Software Development and Management
Took Out Loans
Video Highlights
1. Alex works at Graphistry, a company with a relatively small team, allowing for significant autonomy and collaboration with various team members, including the CEO who is heavily involved in project management.
2. Alex's role involves a wide range of responsibilities, spanning design (Figma), coding (Python and TypeScript), and collaboration with data scientists, highlighting the multifaceted nature of a Product Engineer's work.
3. Alex's background in data engineering is leveraged in his current role, as he develops tools to simplify data analysis tasks, such as SQL query generation, demonstrating the value of a diverse skillset in the field of product engineering.
Transcript
What are your main responsibilities within your current role?
Currently, I work at Graphistry. Our team has around 10 to 15 people, though I'm a little fuzzy on the exact number. Some are contractors, and there's even a subcontractor who manages other contractors, making the day-to-day structure a bit amusing.
I mainly work with about four people, maybe a couple or three others, so roughly eight people in total. The company is a bit funny because our CEO is incredibly productive. One of my coworkers jokes that he works as much as five people, and it's true; he's more productive than anyone I've worked with. It's good to be led by someone so motivated and capable.
In practice, the CEO often acts as the project manager for my work. While I'm a senior engineer with a good degree of autonomy, he's the one tracking our progress towards launch goals, which are typically a couple of weeks away. I have a lot of flexibility within that framework. Since I'm part of a team and trust him, if he emphasizes the importance of launching, I do my best to achieve that.
I also track things he might not have as much visibility into, especially since I'm doing a lot of coding related to our architecture. This involves architectural questions. We're building a product that uses LLMs, like ChatGPT, and we use their API. We'll also have a setup to easily integrate different LLMs in the future.
My daily work involves our app for this, called Louie. The name comes from "LLM UI." When we started, I created Figma designs and sketches. Then, the CEO, our head data scientist, and I collaborated to refine the design. We prototyped it and have developed demo software, which we are now converting into a production application.
This is what I like about the startup environment: you can have a lot of independence and autonomy. You can build things from different aspects, like considering design, various coding challenges, or talking to users. Currently, I'm mostly coding in Python and TypeScript. I've written a significant portion of our application, including the web server and front end.
We're also building out an agent architecture. This involves different AI-enhanced functions that can be run on data in a user-configurable way. An easy initial application is to help users write SQL queries or perform other tasks. My background in data engineering has been useful here. In the past, I wrote queries for YouTube data analytics.
Now, we're creating a tool to help analysts perform these tasks with less effort. Even though I know many programming languages, I don't always remember the specifics of SQL, so I have to re-learn it. This tool is proving helpful in that regard.
I realize I've rambled a bit. Right now, I'm mostly an individual contributor. I do collaborate, for instance, by helping onboard a new team member and discussing strategies for implementing features. Our team is predominantly senior, which is possible because it reduces management requirements due to everyone's high level of autonomy.