My name is Jennifer Ye, and I’m a combined BS/MS computer science major at The Johns Hopkins University. This summer, I am working as a machine learning intern for EcoMap Technologies. Besides being a cat mom, late-night coffee drinker, and someone who always enjoys creating things, I’m also a student entrepreneur at the Spark Accelerator program at the Johns Hopkins Pava Center for Entrepreneurship. As a student entrepreneur, I’ve experienced the challenge of navigating countless funding resources on Google, only to hit a glass ceiling in filtering out the right ones. We live in an information-rich yet noisy digital world where the right information is hidden under layers of hyperlinks. I often find myself spending hours on Reddit or startup forums, ending up frustrated and unsuccessful.
I heard about EcoMap because it was founded by Pava LaPere, the same inspiring Johns Hopkins female entrepreneur after whom the Fast Forward University was re-named. Like my venture, RelationSync, EcoMap is about relationship building, but on a community scale with AI. I’m incredibly grateful for the opportunities the InBaltimore internship program has given me. It has allowed me to dive deeper into the company’s mission and vibrant culture while learning emerging AI technology that will be useful for my own venture.
Company’s Mission:
An ecosystem is a complex network of relationships, resources, and activities within communities. EcoMap believes that every organization has great resources that are often unseen. Our mission is to make these “hidden” resources more accessible, creating more equitable outcomes. Imagine you’re a small-business coffee shop owner in Baltimore. Perhaps you only know one coffee bean supplier that you have been using the whole time. With EcoMap, you can see all the coffee bean options available in the local region and learn about upcoming community events that might need coffee catering or vendors. It’s truly a win-win situation for all stakeholders. Essentially, you can think of EcoMap as a bridge connecting resource providers and seekers, maximizing information utilization. With EcoMap, companies can automate the data scraping and management process and generate actionable insights about their ecosystem.
First Day:
I went in a little nervous since it was my first in-person, computer science related internship and I had no expectations of what it would entail. However, upon arriving at the office, all my stress and uncertainty disappeared. It was early in the morning, and the office was still quiet. The high-ceiling, open office space reminded me of the M-level in Brody Library. Soon, I was greeted with a moving shadow—Stella, one of the resident puppies here in the office. I absolutely loved that it was a pet-friendly environment—literally, you can see 3 or 4 pets around at the same time. Ed, the Vice President of Engineering, chose a seat for me in the center rows by the windows. As everyone started arriving, I was introduced to different people who were truly so friendly and welcoming, and they showed me around the office from top to bottom, including the coffee station and different meeting rooms.
It was Monday that day, and every Monday at noon, we have a company-wide alignment meeting where we sit around a cozy couch and discuss important updates within the firm (engineering feature updates, recent deals, birthday celebrations). One of the best things about working in a startup is that you get full transparency about what the company is up to. Even as an engineer, I know all the updates on the business side, including strategic moves, conferences to attend, and competitor analysis. In other, larger companies, you’d never get that deep of an insight as an engineer.
Daily Schedule:
Every morning, our engineering team has a 30-minute stand-up where everyone reports their project progress and plans for the day. I absolutely love these sessions because they help me quickly learn about everyone’s roles and the exciting projects they’re working on. It also makes me feel heard and seen, reminding me that I’m not just working alone in silos, even if I only interact with two other people on the project. These stand-ups also serve as mini brainstorming sessions where, if someone encounters roadblocks, everyone pitches in ideas and suggests resources from their expertise.
After the stand-up, everyone focuses on their projects for the day. Since I’m working on machine learning tasks, I usually meet with other machine learning engineers on the team to consult on my approach before implementing solutions. I learned the hard way that it’s essential to sort out a high-confidence, high-level approach before diving into implementation. You can often find me in the “cozy corner” meeting room for deep focus sessions or standing next to a whiteboard breaking down problems with other engineers.
As an unfortunate victim of food coma and to adapt to a corporate no-napping lifestyle, I have very light lunches (sometimes ordered by the company—thank you, EcoMap). As the afternoon progresses, I like to grab coffee with Valentine, another ML engineer on the team. We’ve bonded so much on our coffee walks, which are one of my favorite things about EcoMap. We love Café Los Sueños (this blog post was actually written there) and its owner, and enjoy chatting with him about everything from his new summer ginger drink to his hometown while waiting for our coffee.
My Work:
I’m working on the machine learning/data science side of things, focusing on three projects. My main project is the Keyword Revamp project. This involves tagging keywords in resources for information extraction and filtering purposes. Recently, we transitioned from manual keyword generation to a fully AI-automated process. However, our initial AI model is still in its early stages and often produces contradictory or redundant keywords. My task is to develop a robust, permanent testing scheme to benchmark the performance of future ML models. This involves quantifying “keyword tagging performance” using three methods: text similarities, semantic similarities, and run consistency. During the implementation phase, I experimented with various text preprocessing methods and ML models to achieve the most unbiased scoring.
One of my side projects is the Data Export project. While most of our older clients receive a website as the deliverable, some of our newest clients have requested data-only deliverables. My task is to research state-of-the-art storage options, including Amazon S3 and Google Cloud Storage, and streamline the data format for export.
Another side project I am working on is the Accessibility Project. The goal of this project is to make our website navigable for users with learning or visual disabilities using screen readers. My task was to add over 200 header tags to more than 10 pages of our website.
Things I Love About EcoMap:
The mentorship at EcoMap has been invaluable; everyone in the company has acted as a mentor to me at some point. When I encounter a programming crisis or hit a dead-end, someone always takes time out of their busy schedule to sit with me and work through the problem together. Even during lunch, the business director readily shares stories about competitors or recent deals if I’m curious. I know for sure that people at EcoMap have my back. Despite some tasks being explorative or ambiguous, I’ve never felt thrown into a clueless task.
I also appreciate the transparency at EcoMap. People unapologetically ask harsh but honest questions or voice concerns when needed for the company’s growth. It feels liberating here at EcoMap, where everyone’s voice is heard and work is seen. I personally love knowing what everyone is up to, and collaboration is evident in every task. For instance, if someone escalates a problem on our webpage at 10 PM during a holiday, you can see at least four people immediately jump onto the Slack channel to solve it.
Final Words:
This has been one of my absolute favorite summer internships. Before this internship, I switched from being a pre-med student to computer science major, resulting in imposter syndrome and doubting I was enough for an ML engineer role, especially since I barely knew machine learning. This experience reassures me that it is never too late to switch paths (ML is not as monstrous as people describe it to be). The people at EcoMap have been incredibly accommodating and patient in teaching me new things. They are also super knowledgeable, and almost every question I raise has an answer. This job has made me look forward to every day this summer, and I can already imagine how emotional I will be when walking into that brick building is no longer part of my daily routine. So thank you, EcoMap, and the InBaltimore program for this amazing experience.
