timeline
2022 - 2024
I joined an AI startup, Aethos Labs, to build machine learning solutions for sales industry. My work includes data gathering, cleaning and preparation, and experimenting with NLP models in order to predict engagement of prospects with sales campaigns, and other tasks such as sentiment analysis, deduplication, and detecting bots.
2021 - 2022
I decided to apply for a mentorship program at SharpestMinds, where I developend an NLP application for researchers that elevates their workflow with summarization and classification of scientific articles, and title generation based on an abstract.
2017 - 2020
PhD at Polytechnique Montreal, on developing algorithms for highly nonlinear structures that deform under fluid flow. My advisors were Eric Laurendeu and Frederick Gosselin.
I wrote my findings in Flutter Instability of a Reconfiguring Beam with Large Displacement and decided to drop out from PhD and instead join tech industry.
2009 - 2016
MSc, BSc at Sharif University of Technology where I worked with Masoud Darbandi on simulation of engineering systems, e.g. aerodynamics of wind turbines and combustion phenomena in industrial incinerators.
featured writing
Send scheduled messages from a bot to a channel in Microsoft Teams
Why design is so hard, with examples (soon!)
It's mostly the same thing, but with different names (soon!)
Maybe you should think deeply on career path (soon!)
selected projects
It's sort of surreal to see an AI agent that decides to act on itself to reach some goals!
I believe that it's the most exciting avenue for large language models, to act as the brain of a system that has a set of tools at its disposal. I am going to experiment with this idea here.
The deep learning revolution was attracting widespread attention, and I thought that many tasks for researchers can be done by AI. So I made a simple AI-powered assistant to do basic tasks for the people in science. This was way before (two years) ChatGPT era.
Navier–Stokes existence and smoothness is one of the Millennium Prize Problems with a US$1 million prize!
Although I did not try to solve this problem (and probably you shouldn't unless you're a math genius), I wrote a computational fluid dynamics code to simulate the nonlinear reconfiguration of a beam in the fluid flow.
I can't share the code since it's proprietary (sorry!) but some cool results are in my thesis.
1D piston-spring system was one of my toy projects in my PhD days, inspired by super interesting work of Frederic J. Blom. I was trying to understand the interaction of two nonlinear systems.