Work experience

Head of Infrastructure and Security at Medeloop
Palo Alto, CA, Jun 2022 – Present
  • Leading client and API deployment on AWS and organizational security posture for seed-stage rare disease data platform startup.
Software Engineering Intern, Security and Privacy at Verkada
San Mateo, CA, Jun 2023 – Sep 2023
  • Established automated firmware and network security testing program and Linux hardening standards for physical security devices
  • Test implementation and enforcement substantially reduced device attack surface and improved security and compliance posture
Software Engineering Intern at Lacework
San Jose, CA (remote), Jun 2022 – Sep 2022
  • Engineered end-to-end virtualization of benchmarking system on Spark, reducing data import time by 20x vs. Snowflake.
  • Contributed enhanced Snowflake and Spark parsing support to SQLGlot, an open-source SQL parser and transpiler. 3 PRs merged.
Engineering Intern at Uptycs
Waltham, MA (remote), Nov 2020 – Apr 2021
  • Wrote and deployed production feature to Osquery monitoring software to inspect and detect malware in Java packages.
  • Functionality used to detect and patch client software with Log4Shell vulnerabilities (10.0 CVSSv3 base score CVE).
  • Code now open-source.
Security consultant for early-stage startups
Jun 2022 – Present
  • Evaluating and strengthening initial setup and ongoing security of tech stack (incl. Firebase, AWS) for Stanford startups.
Research Science Institute Intern at Akamai
Cambridge, MA, Jun 2019 – Aug 2019
  • Engineered realtime garbage collection monitoring system for Go programs with per-thread granularity.
  • Detailed flagging of stop-the-world pauses used for profiling and boosting performance across Akamai Labs codebase.

Teaching, leadership, and competition experience

President and CCDC Linux & Cloud Lead at Stanford Applied Cyber
Stanford, CA, Jan 2021 – Present
  • Securing Linux and AWS systems against external red team in CCDC competition environment.
  • Member of the 2023 National CCDC Championship (1st place) Stanford team.
  • 3rd place finish at National CCDC 2022; 1st place at Western Regional CCDC 2022 and 2023.
  • Lead security basics workshops for beginners (Apr ‘22, Oct ‘22) and application security workshops for entrepreneurs (Jan ‘23).
  • Found and disclosed web, mobile, and cloud security vulnerabilities to 10+ Stanford student startups, leading to fixes to protect sensitive personal data. Work covered in the Stanford Daily.
  • Presented on vulnerabilities in GraphQL client apps (Mar ‘23) and Gradescope autograders (Apr ‘23), influencing autograder design for Stanford CS courses.
  • Presented on vuln-finding in Firebase apps (Feb ‘22) and working around Google OAuth for security research (May ‘22).
  • Contributed Google OAuth sign-in support to Baserunner, an open-source Firebase exploration tool.
Teaching Assistant and Infrastructure Lead at Stanford University (via Stanford Internet Observatory)
for INTLPOL 268 Hack Lab taught by Alex Stamos and Riana Pfefferkorn, Sep 2022 – Dec 2022
  • Hack Lab is Stanford’s intro cybersecurity, cyberlaw, and cyber policy class, with 170+ enrolled students.
  • Taught two discussion sections (44 students),
  • Designed and implemented course cloud infrastructure on GCP at scale.
  • Designed and implemented several new lab exercises, including on cracking an encrypted WiFi packet capture, breaking into Windows machines with EternalBlue and psexec, and leaking data from an insecure Firebase mock chat app.
  • Responded to an incident involving an EternalBlue attack on course infrastructure.

Research experience

Software patching dynamics (Stanford ESRG, Oct 2022 – Present)

  • Exploring how different organizations and enterprises patch software security vulnerabilities over time on the open internet, and on how and when attackers target those vulnerabilities.
  • Advised by Zakir Durumeric.

Parallel, human-interpretable machine learning (Harvard/UBC/Duke, Jun 2017 – Feb 2022)

  • Significant contributor and co-author to work and papers on Certifiably Optimal Rule Lists (CORELS), a system to generate human-interpretable machine learning models for tasks such as crime recidivism prediction.
  • Key contributor to parallel implementation of the algorithm, achieving linear speedup and increasing tractability on 250k+ sample datasets.
  • Co-wrote short paper on systems optimizations presented at SysML Conference (now MLSys) 2018.
  • Created NodeJS web UI and R API.
  • Advised by Margo Seltzer and Cynthia Rudin.

Rust concurrency evaluation (MIT PDOS, Jan 2018 – Apr 2019)

Adversarial machine learning (MIT Madry Lab, Jan 2019 – Jun 2020)

  • Explored effectiveness of ensembling with robust and non-robust features for robustness under adversarial attack.
  • Developed ensemble schemes that yield same adversarial robustness as a single model but improve natural accuracy.
  • Paper published in ICLR 2020 workshop on trustworthy machine learning (44% acceptance rate).
  • Advised by Aleksandr Madry.

Virtual assistants for customer support queries (Stanford OVAL, Apr 2021 – Dec 2021)

  • Created virtual assistant pipeline to classify customer support requests with GPT-3 data augmentation; increased sample data size by 4x.
  • Advised by Monica Lam.

Open-source projects and contributions

Owned projects

Contributions

Publications

Aditya Saligrama, Guillaume Leclerc. Revisiting Ensembles in an Adversarial Context: Improving Natural Accuracy. ICLR 2020 Workshop on Towards Trustworthy ML: Rethinking Security and Privacy for ML. Presented April 26, 2020.

Aditya Saligrama. KnowBias: Detecting Political Polarity in Long Text Content. AAAI 2020 Student Abstract and Poster Program. Presented February 9, 2020.

Aditya Saligrama. KnowBias: A Novel AI Method to Detect Polarity in Online Content. CoRR, 2019.

Aditya Saligrama, Andrew Shen, Jon Gjengset. A Practical Analysis of Rust’s Concurrency Story. CoRR, 2019.

Nicholas Larus-Stone, Elaine Angelino, Daniel Alabi, Margo Seltzer, Vassilios Kaxiras, Aditya Saligrama, Cynthia Rudin. Systems Optimizations for Learning Certifiably Optimal Rule Lists. SysML (now MLSys) Conference, 2018.

Selected long-form blog posts

Note: these posts were selected due to their exploration and synthesis of existing and new security vulnerabilities or techniques, providing a guide for future related security work.

Talks

Education

M.S. candidate in Computer Science, Computer and Network Security track
at Stanford University, Feb 2023 – Jun 2025 (expected)
  • CS 255 Introduction to Cryptography (W ‘22)
  • CS 251 Cryptocurrencies and Blockchain Technologies (F ‘22)
  • CS 249I The Modern Internet (W ‘23)
  • CS 229 Machine Learning (F ‘21)
  • CS 155 Computer and Network Security (S ‘22)
  • CS 153 Applied Security at Scale (W ‘23)
  • CS 144 Introduction to Computer Networking (S ‘23)
  • INTLPOL 268 Hack Lab: Introduction to Cybersecurity (F ‘21, TA F ‘22)
  • CS 199 Independent Work – Research with ESRG (F ‘22, W ‘23)
B.S. candidate in Computer Science, Systems track
at Stanford University, Sep 2020 – Jun 2024 (expected)
  • CS 224U Natural Language Understanding (S ‘21)
  • CS 191W Senior Project – Research with ESRG (S ‘23)
  • CS 161 Design and Analysis of Algorithms (W ‘22)
  • CS 154 Introduction to the Theory of Computer Science (F ‘21)
  • CS 152 Trust and Safety Engineering (S ‘23)
  • CS 149 Parallel Computing (F ‘22)
  • CS 143 Compilers (S ‘22)
  • CS 140E Operating Systems Design and Implementation (W ‘22)
  • CS 110L Safety in Systems Programming (S ‘21)
  • MATH 104 Applied Matrix Theory (S ‘21)
  • DESIGN 151 Designing your Business (S ‘23)
  • LINGUIST 150 Language and Society (W ‘23)
  • PSYCH 1 Introduction to Psychology (F ‘22)
  • CLASSICS 136 The Greek Invention of Mathematics (S ‘22)
  • MI 70Q Photographing Nature (W ‘22)
  • MUSIC 2C Introduction to Opera (F ‘21)
  • ECON 50 Economic Analysis I (W ‘21)
  • SOC 9N 2020 Election – Introsem (F ‘20)