Week 5 & 6
These two weeks have been steady but eventful: balancing classes, career events, and learning more about some really interesting startups.
Startups2Students
One of the highlights was getting to learn about startups that are applying AI in surprising ways:
4 Ever Oceans is building a global community for ocean enthusiasts. They’re not only sharing ocean photography, but also using AI for marine life identification and creating tools for education, research, and even a marketplace. I liked how approachable it felt: you don’t need to be a scientist to contribute, and yet it still supports real research.
Zooly Labs stood out with its bold idea: replacing $2M surgical robots with $750 smart sensors. Their combination of AI, sensors, and real-time algorithms aims to make surgeries more efficient and affordable. It felt ambitious, but also practical in a way that could have real impact in healthcare.

- Faex Health takes a completely different approach by focusing on gut health. Using deep learning, fingerprinting, and color density analysis, they’ve developed a system to detect diseases from stool images with over 99% accuracy. It’s the kind of application that could make a big difference in diagnostics.
Rippey AI is tackling something less visible but equally impactful: email and document automation in logistics and supply chains. Their AI-powered response bots are designed to cut down the time spent on repetitive communication tasks. It’s not the kind of idea that grabs headlines, but it’s practical and directly useful for industries where time and accuracy really matter.
Hearing these back-to-back made me realize how wide the scope of AI is. These projects showed that innovation doesn’t always come from expected places.
It also made me think about CU Boulder’s culture: the university was recently ranked No. 1 in the nation for launching startups based on university discoveries. Sitting in a room full of founders, I could see why. There’s a real energy here around turning research into impact, and it’s motivating to feel part of that environment.
More Events On Campus
I also made time for a few CU events that were simple but valuable.
International Coffee Hour was exactly what it sounded like: coffee, snacks, and conversations. It was a nice pause from the usual routine and a chance to meet people from different parts of the world.
The Career Services Open House Extravaganza gave me the chance to finally get some professional photos, meet the team, and pick up some useful advice about navigating career fairs and workshops. Sometimes those small steps feel like progress, too.
The Professional Experiences Showcase was bigger and more packed. It highlighted opportunities like internships abroad, volunteering, mentorship, and research. It felt more like a reminder of just how many directions there are to grow in, and that there’s no shortage of opportunities to explore.

The Databricks MDM workshop was another highlight. LakeFusion, an MDM solution built directly into the Databricks Lakehouse, showed how entity resolution, data standardization, and creating “golden records” can now happen inside the platform itself. What I liked most was how it removes the need for separate MDM systems: the work happens where the data already lives. It feels like a natural step forward in making data management more seamless and impactful.
Another important session I attended was the Post-Completion OPT Employment Workshop hosted by International Student & Scholar Services. As an international student, this one really mattered. The workshop broke down the OPT process, from application timelines to maintaining status while working, and cleared up a lot of small questions I had. It wasn’t just about the paperwork, but also about understanding how to plan ahead and make sure I’m set up for the kind of opportunities I want to pursue after graduation.
Coursework
Classes kept me busy and balanced. Each one brought something different:
NLP: We covered n-grams, smoothing, sentiment analysis, and spam detection. A lot of time was spent on evaluation (things like perplexity and entropy), which made me realize just how important it is to measure models carefully.
Neural Networks and Deep Learning: This class had me coding logistic regression by hand. Binary Cross-Entropy, sigmoid activation, backpropagation: all of it step by step. It was time-consuming, but it helped me see the mechanics more clearly than if I had just used a library.
Information Visualization: Reading Storytelling with Data by Cole Nussbaumer Knaflic made me notice how small design choices: gridlines, orientation, colors change the way people understand a chart. It reinforced that visuals are about communication, not decoration.
These two weeks reminded me that learning happens in different settings in class, at career events, and in conversations with people working on unusual ideas. It doesn’t always feel fast-paced, but steady progress has its own value. I’m glad I made the effort to show up for these opportunities, even the small ones, because they add up.
Comments
Post a Comment