Subscribe to Impira's blog
Stay up to date with all things Impira, automation, document processing, and industry best practices.
What’s your official role here at Impira?
I’m a software engineer on the Machine Learning (ML) team.
Describe what you do in a few sentences.
I solve fun problems. Lately, there are two I’ve been deep in. The first is document classification — can we organize and curate documents the way you would? The second is entity extraction — can we excel at helping you find important information from your documents, combining methods from natural language processing (like BERT and friends) with our causal modeling approach?
How does your work impact people’s lives?
Just two months after I joined Impira, our wonderful Chief Product Officer (Lorilyn) slid me into the DM’s – i.e., introduced me to one of our customers over Slack. Our customer described how cumbersome it was for them to have to manually sort their documents into the right collections. They asked if it was possible for us to automate this, and it was! While organizing a pile of documents may seem like a small task, being able to remove real points of friction for real people makes my inner machine hum.
Impira’s mission is to make data entry a thing of the past. What does this mean to you?
People create documents for other people to read, not for computers to read, and rightly so. However, because our computers are unable to read, people are often stuck playing interpreter for them – personally, I’ve lost count of how many times I’ve had to manually, painfully transfigure restaurant receipts into expense reports, watch my doctor’s receptionist enter in my medical history, and sort through bank statements to do my taxes.
But, if computers can interpret code, what’s stopping them from interpreting documents as well ? Impira will make documents accessible to computers, and as a result, make computers more humane.
Tell us about the culture of inclusion at Impira.
To me, inclusion implies honesty delivered with whole-hearted kindness and good intent. Often, honesty and kindness are viewed as dichotomous things, but I don’t think they need be. At Impira, we are able to be rigorous and thoughtful in our work because we deeply trust that we care about each other and the people we serve. I don’t think it would be possible to build a great product without this level of mutual respect.
What do you love most about working at Impira?
We really care. About the product we’re building and the people we are building it with. At the end of the day, machine learning is a technology (albeit a powerful one) to accomplish human goals.
Impira is creating the future of machine learning. Why should people come build with us?
We value our craft. We consciously prioritize deep work, precisely articulate our goals, and choose to do one thing well over many things half-heartedly.
We solve technical problems from first principles and end-to-end. On the ML team, we have a diverse skillset – from Bayesian modeling to computer vision to natural processing to runtime optimization – which we creatively blend. On top of that, we also partner closely with the Crucible (infrastructure) and Core Product teams and learn greatly from their expertise and feedback.
Last but certainly not least, we throw pretty great off sites!
Describe your journey into machine learning. How did you choose that career?
I wandered into machine learning as an undergraduate at the University of Washington. I was working on an interdisciplinary research project at the intersection of systems and human-computer interaction (HCI), using crowdsourcing to improve speech recognition quality for Hindi and low-resource Indian langauges. That work led me to explore natural language processing (NLP). At that time, deep learning also started taking off and completely rewriting everything we knew about how to do NLP. I became curious and the rest is history.
What made you want to work at Impira?
I actually hadn’t been planning on leaving my last job, but when I saw a demo of Impira, I was immediately intrigued. Impira wasn’t just talking about facets of cutting-edge ML — incremental execution, human-in-the-loop feedback, one-shot learning — but had already built them into the product, and beautifully well, with such a small team at that.
While interesting technical challenges brought me through the door, the people are hands-down what convinced me to call Impira home. I’m proud and humbled to call every individual at Impira my colleague.
What are some of your passions outside of Impira?
I love long-distance running, attempting new vegan recipes, and listening to audiobooks.