Shefali Ramakrishna Westfield

PhD candidate in Operations Research, graduating May 2027. On the job market. Research in queueing theory, scheduling, and stochastic decision-making under uncertainty.

Email: sr899@cornell.edu

CV LinkedIn arXiv Blog
← Back to blog

Building an Anonymous Q&A Tool for a 150-Person Lecture

Before this semester began, I wanted a way for students to respond anonymously to questions in real time during lecture. I built a small web app: students submit answers on their phones, I approve and display them on screen, and I can talk through common misconceptions with the whole class without singling anyone out.

I spent a weekend building this: a Firebase Database on the backend, with a Cloud Function to summarize if there were too many questions (at the suggestion of my father-in-law, a former teacher. thanks, Daryl!). And then, a few weeks in, I discovered my co-instructor had Poll Everywhere access through the university the entire time. Oops.

Would I build it again knowing that? Almost surely yes!! The instinct to build it in the first place says something about how seriously I take the question of feedback, and it means I'm not dependent on a specific tool being available to run a large course well.

What I actually care about isn't the tool itself, it's the underlying problem: in a lecture of 150 students, relying only on a hand raised in the air to gauge confusion is a terrible signal. That requires someone is confident enough to ``interrupt'' class... in front of everyone... admitting there's a concept they don't yet understand. The students who most need to ask a question are usually the ones least likely to raise a hand (I know this from personal experience!). An anonymous channel, whether it's something I built or something the university already licenses, closes that gap.

Now I use Poll Everywhere day to day, since there's no reason to maintain my own infrastructure when a supported tool does the job. But building it first taught me a lot about what I actually wanted the tool to do, which was a useful exercise in itself.

Picture of Shefali with ``Latte'', a Spectacled Owl.

About Me

Hi! My name is Shefali (she/her). I am a fifth-year PhD candidate in Operations Research at Cornell University, working with Ziv Scully on scheduling in queueing systems. I am currently instructor of record for ORIE 3500.

I have a bachelor's and master's degree in mathematics with honors from Bryn Mawr College, where I graduated magna cum laude.

When I’m not thinking about research "queue-stions", I enjoy birdwatching (ask me about American Robins and American Crows!), embroidery (particularly of birds), improv, turn-based tactics and strategy games, deckbuilding roguelikes, and Shakespeare plays!