Hi there! Welcome to my homepage; I go by Sruti (pronounced as /ʃrut̪i/ or sh-ru-thi).
I am a human-computer interaction & software-engineering research. I am interested in helping programmers program effectively and efficiently. This includes helping them learn to program, author programs, seeking help, debugging, or even versioning.
Currently, I am a postdoctoral researcher in the Calc Intelligence group at Microsoft Research, Cambridge, United Kingdom. My current research aims to make programming easy and error-free for citizen programmers and data analysts, with little to no programming training, but still having to program in some form. Such people who program to an end (e.g., accountants writing spreadsheet formulas) are called end-user programmers, and the activity itself is called end-user programming.
Before joining Microsoft Research, I got my Ph.D. from Oregon State University, USA, advised by the awesome Margaret Burnett. My dissertation titled “Variations Foraging” is all about the theoretical underpinnings of how people reason about variants in their head, and how that influences the way they work with them. Variants, here, mean different copies of the same artifact; it includes versions but is not limited to them. My dissertation established that a successful HCI theory, namely the Information Foraging Theory, also applied to the case of information foraging among very similar variants.
What is so surprising about it, you ask?
Information Foraging Theory posits that people sniff their way to their information prey, just like predatory animals in the wild sniff their way to their prey. Translated to a concrete example, if I am looking for a figure from my thesis, then I would go to my computer, look for a file that reads something like thesis, or dissertation, or Variations Foraging, or even [Ragavan,2019] to open the file and pick search for the figure in there. But what happens if I want to restore a figure that I deleted from my final version of the thesis, but I created and had as part of one of the previous revisions. Worse, what if the files are named thesis-1.docx, thesis-2.docx, …. and so on. The “scent” from all these filenames are very, very similar and it is hard to guess which of them might hold the figure I am looking for.
Now, Information Foraging Theory had not dealt with such cases of similar variants of an artifact, but has only largely focused on foraging among largely dissimilar information. A question remained: does the notion of scent exist even when the scent is all very, very similar to each other? If not, how would people forage in such scenarios. Using a series of studies, and computational models, my dissertation established that IFT does explain information foraging among similar models. Turns out, in such cases, the foraging strategies are very different, and are all about various forms of comparison.
Practically, if you think your environment is going to have variants, make differences and similarities obvious. Build in cues that exactly tell what is different or similar, quickly. For version control, capture what behavior remains / goes away with each version, in addition to changes in code. For changes in UI, capture the UI changes as screenshots / images, so people can quickly tell what has changed where.
Along the way, I also got a Masters from Oregon State. My MS Thesis was all about what software developers do with their version control tools — how they commit their changes, what information they seek, what problems they face, and what information foraging theory has to say about designing better version control tools.
Both my Masters and PhD theses research have won Best Paper Awards at top HCI and SE conferences.
Before grad school, I was a software developer at ThoughtWorks. I still see myself as a developer and actively write code (e.g., build my own prototypes, analyze data, or even for fun personal things).
When I am not doing something related to Computer Science, I am reading on Indian history, philosophy, textiles, arts and mythology. I’m also a Sanskrit teacher and student. My recent fun activity is the study of Panini’s Ashtadhyayi (in bits and blobs) and Valmiki Ramayana in its original, along with its Sanskrit commentaries. They’re both mind-blowing in terms of sheer meticulousness (no unsubstantiated claims, discussion of alternative claims and exceptions, mostly unambiguous, only working on parts that are novel and need clarifications, lack of redundancy, brevity and all that we think is a hallmark of good research writing).
When I’m not doing any of this, I love bothering my husband, giving close friends and family a hard time, taking long walks, daydreaming, writing and doing Nothing.
- I read, write and speak two ancient classical languages, Tamil and Sanskrit.
- Starting Fall 2021, I am a high school teacher in the US, teaching Sanskrit for foreign language credits via the SAFL program.
I’m married to Dr. Sudarshan Narayanan, a materials scientist, violinist and Sanskrit teacher.