hi, i'm arlene siswanto

iā€™m an MEng (fall '20) / BS (spring '20) studying computer science and artificial intelligence at MIT. my areas of interest include machine learning, math, software, startups, and design.

i enjoy
developing, designing, creating, building, and learning
cool things and hope to work on
meaningful projects with great impact

experience

coursework at MIT // graduate level: 6.864 natural language processing, 6.435 bayesian modeling, 6.883 machine learning, 6.819 computer vision, 6.857 computer and network security, 6.S974 decentralized applications, 6.840 theory of computation. mathematics: 18.650 statistics, 18.615 stochastic processes, 18.600 probability, 18.065 matrix methods

late 2020 - 2021 // (incoming) ai resident at google brain, new york. will conduct deep learning research with the brain team during 12-month residency.

spring 2020 // researcher at mit csail, cambridge. explored the lottery ticket hypothesis and its relation to batch normalization in neural networks.

fall 2019 // research platform intern at deepmind, london. developed alongside research platform team during fall semester on platform that allows researchers to perform an initial analysis of their experiments.

summer 2019 // software engineering intern at jump trading, chicago. developed on equities trading team to automate configuration verification through recursive AST traversal. built standalone trading microplatform with real-time connections to exchanges. c++, python.

summer 2018 // software engineering intern at bloomberg, new york city. developed on the emsx trade execution management platform in the trading and analytics department. c++, javascript, python.

summer 2017 // software engineering intern at truemotion, boston. implemented a data visualization platform to allow data scientists to better visualize their data. angular, javascript, html, css.

projects

time series forecasting, 2020 // implemented lstm, arima, holt-winters, and prophet models to analyze forecasting abilities when applied to time series data. paper

dormsp.am, 2019 // platform built for the MIT community that scrapes, parses, and aggregates oncampus events advertised on email. website

image colorizer, 2018 // given a black-and-white image, returns a colored version of the image. we first input the image through an image classifier, then through a convolutional neural network trained on colored images from a specific category. paper

beaverdocs, 2018 // decentralized google docs. using a conflict-free replicated data type for quick insert and delete, as well as a peer-to-peer broadcasting system, we created a platform to allow multiple users to edit the same document without a central server storing the contents. website, paper

diff, 2018 // open platform that allows researchers, data scientists, and holders of data to share potentially confidential datasets in a differentially private manner. website, paper

annotateAR, 2018 // hackprinceton ā€™18 best ar/vr prize. a real-time, collaborative augmented reality platform that allows users to annotate the physical world. devpost

meter, 2017 // platform that allows owners of privately-owned parking spots to lend their spots out to drivers seeking cheaper and more convenient parking. received mit sandbox startup funding. project

pillAR, 2017 // pennapps ā€™17 second place winner. augmented reality application that tracks the number of pills taken and scrapes relevant medication information. devpost

have any projects in mind?

email me at siswanto (at) mit (dot) edu
connect with me on linkedin