Abhinav Aggarwal
I am a Senior Applied Scientist at Amazon Inc, working in Seattle, WA for the AWS Marketing Technologies team on next-generation marketing content creation tools using generative AI technologies. Before this, I worked with Amazon Search Science and AI org on responsible Artificial General Intelligence (AGI) research, specifically within search content moderation, trust, safety and privacy for Large Language Models (LLMs). I have also worked with Amazon Trust and Privacy organization with the Devices and Services vertical on privacy and trust enhancing technologies for enhancing builder experience within the Amazon smart home device ecosystem. My work is mainly at the intersection of Trustworthy AI and product development. I also help with risk management with respect to privacy regulatory compliance and have experience with several cross-org projects on incidence response, optimizing data availability for builders, and taking business-driven trusted AI decisions for our key stakeholders. Prior to this, I worked with Cornell University, VISA research, Google, and Microsoft for internships spanning secure computing, software development, distributed ledger technologies (blockchains), and NLP-driven privacy. I obtained a Ph.D. in Computer Science in 2019 from the University of New Mexico (UNM), Albuquerque, NM under the supervision of Prof. Jared Saia.
List of Publications
Privacy-Preserving Machine Learning
Reconstructing Test Labels from Noisy Loss Scores (Full paper). [AISTATS 2022]
Aggarwal, A., Kasiviswanathan, S. P., Xu, Z., Feyisetan, O., Teissier, N. [pdf]Balancing Utility and Scalability in Metric Differential Privacy. [UAI 2022]
Imola, J., Kasiviswanathan, S. P., White, S., Aggarwal, A., Teissier, N [pdf]Reconstructing Test Labels from Noisy Loss Scores (Extended Abstract). [NeurIPS 2021]
Aggarwal, A., Kasiviswanathan, S. P., Xu, Z., Feyisetan, O., Teissier, N. [pdf]Label Inference Attacks from Log-Loss Scores. [ICML 2021]
Aggarwal, A., Kasiviswanathan, S. P., Xu, Z., Feyisetan, O., Teissier, N. [pdf] [PMLR]On a Utilitarian Approach to Privacy-Preserving Text Generation. [NAACL-HLT 2021]
Xu, Z., Aggarwal, A., Feyisetan, O., Teissier, N. [pdf]Density-Aware Differentially Private Textual Perturbations Using Truncated Gumbel Noise [FLAIRS 2021] — Best Paper Award!
Xu, N., Feyisetan, O., Aggarwal, A., Zu, X., Teissier, N. [pdf]Research Challenges in Designing Differentially Private Text Generation Mechanisms. [FLAIRS 2021]
Feyisetan O., Aggarwal, A., Xu, Z., Teissier, N. [pdf]On Primes, Log-Loss Scores and (No) Privacy. [EMNLP 2020]
Aggarwal, A., Xu, Z., Feyisetan, O., Teissier, N. [pdf]A Differentially Private Text Perturbation Method using a Regularized Mahalanobis Metric. [EMNLP 2020]
Xu, Z., Aggarwal, A., Feyisetan, O., Teissier, N. [pdf]
Bio-Inspired Computation
Ants on a Plane. [SIROCCO’20] — Best Paper Award!
Aggarwal, A., Saia, J. [pdf]LOCUS: A Multi-Robot Loss-Tolerant Algorithm for Surveying Volcanic Plumes. [IRC’20]
Erickson, J., Aggarwal, A., Fricke, G.M., Moses, M. [pdf]Ignorance is Not Bliss: An Analysis of Central-Place Foraging Algorithms. [IROS’19]
Aggarwal, A., Gupta, D., Fricke, G.M., Vining, W., Moses, M. [pdf]On the Minimal Set of Inputs Required for Efficient Neuro-Evolved Foraging. [BDA@ PODC'19]
Erickson, J., Aggarwal, A., Moses, M.A Most Irrational Foraging Algorithm. [BDA@ PODC'19]
Aggarwal, A., Vining, W., Gupta, D., Saia, J., Moses, M.On Site Fidelity and the Price of Ignorance in Swarm Robotic Central Place Foraging Algorithms. [PODC’19]
Aggarwal, A., Gupta, D., Fricke, G.M., Moses, M.
Distributed Systems
A Scalable Algorithm for Multiparty Interactive Communication with Private Channels. [ICDCN’20] — Best Paper Award Nominee.
Aggarwal, A., Dani, V., Hayes, Thomas P., Saia, J. [pdf]Thwarting Adversaries with Randomness and Irrationality. [Ph.D. thesis, 2019]
Bootstrapping Blockchains without a Trusted Setup. [PODC’19]
Aggarwal, A., Mohavedi, M., Saia, J., Zamani, M.Multiparty Interactive Communication with Private Channels. [PODC’19]
Aggarwal, A., Dani, V., Hayes, Thomas P., Saia, J.Sending a Message with Unknown Noise. [ICDCN’18]
Aggarwal, A., Dani, V., Hayes, Thomas P., Saia, J. [pdf]
Miscellaneous
Aggarwal, A., Sinha, R., Christodorescu, M. Oblivious Filtering of Data Streams. Technical Report, VISA Research. (2019)
Aggarwal, A., Zamani, M., Christodorescu, M. Revisiting Remote-Attack Prevention: Challenges in Achieving Complete Decentralization. (Technical Report) (2018)
Aggarwal, A., Zamani, M., Christodorescu, M. Incentive-Driven Remote Configuration of Security Gateways. (arXiv: 1709.04569) (2017)
Aggarwal, A. On the Equivalence of Probabilistic Automata: A Brief Literature Review. Technical Report. UNM 2015.
Aggarwal, A. Evolving Random Sequences using a Genetic Algorithm. Technical Report. UNM 2015.
Aggarwal, A. Probabilistic Programs and their Verification: A Brief Survey. Technical Report. UNM 2014.
Aggarwal, A., A Similarity Criterion for Sequential Programs using Truth-Preserving Partial Functions. (arXiv: 1708.05640) Master’s Thesis (2014).
Aggarwal, A., Kumar, P. A Theory of Function-Induced Orders to Study Recursion Termination. Technical Report. IITR 2013.
Aggarwal, A., Srivastava, R., Malik, S., Meena, K., Poonam. Virtual Differential Storage Based k-Rollback Concurrency Control Algorithm in Distributed Shared Memory Systems. Technical Report. IITR 2012.
Jaya, Malik, S., Aggarwal, A., Sardana, A. Novel Authentication System using Visual Cryptography. [WICT'11]
NOTE: In theoretical computer science, it is standard to list author names in alphabetical order.
Patents
Machine Learning
Freeman, C., Aggarwal, A. Locally Faithful Ensembles of Model Agnostic Interpretations. (Submitted) Patent, Verint Intelligent Self-Service (2019).
Freeman, C., Aggarwal, A., Merriman, J., Beaver, I. Model-Agnostic Visualizations using Linear Programming Approximation. Patent, Verint Intelligent Self-Service (2019). US Patent App. 16/356,091
Beaver, I., Freeman, C., Merriman, J., Aggarwal, A. System and Method of Highlighting Influential Samples in Sequential Analysis. Patent, Verint Intelligent Self-Service (2019). US Patent App. 16/283,135
System Security
Aggarwal, A., Zamani. M, Christodorescu, M. Remote Configuration of Security Gateways. Patent, VISA Research (2020). US Patent App. 16/626,478
Ren, J., Aggarwal, A., Salajegheh, M. System, Method, and Computer Program Product for Maintaining User Privacy in Advertisement Networks. Patent, VISA Research (2020). US Patent App. 16/558,910
Aggarwal, A., Sinha, R., Christodorescu, M. On Oblivious Filtering of Data Streams. (Submitted) Patent, VISA Research (2020).
Aggarwal, A., Zamani. M. Sybil-Resistant Identity Generation. (Submitted) Patent, VISA Research (2018).
Work Experience
Full Time Positions
Applied Scientist. Amazon.com Inc., Seattle, WA. (Sept 2019 - Present)
Graduate Research Assistant. Department of Computer Science, University of New Mexico, Albuquerque, NM. (Aug 2014 - Jun 2019)
Co-op/Internships
Research Intern: VISA Research, Palo Alto, California, USA, Summer 2018 (Oblivious data filtering. Abhinav Aggarwal, Mihai Christodorescu, Rohit Sinha)
Visiting Researcher: Cornell University, Ithaca, New York, USA, Fall 2017 (A Folklore reduction from State Machine Replication to Binary Agreement under Partial Synchrony. Abhinav Aggarwal, Elaine Shi)
Research Intern: VISA Research, Palo Alto, California, USA, Summer 2017 (Remote attack prevention and PKI establishment. Abhinav Aggarwal, Mihai Christodorescu, Mahdi Zamani)
Software Engineering Intern: Google Inc., Mountain View, California, USA, Summer 2016 (Provisioning an Isolation Manager and Studying the Feasibility of Using Linux Control Groups for User-Level Isolation in Search Queries. Abhinav Aggarwal, Rahul Bansal, Chris Bond)
Software Engineering Intern: Microsoft India Development Centre, Hyderabad, India, Summer 2013 (Analyzing Test Frameworks and Dependencies for Porting the PDF-API test code. Abhinav Aggarwal, Ravindra Sushma Bhartiya)
Software Engineering Intern: Microsoft India Development Centre, Hyderabad, India, Summer 2012 (Optimizing VHD (Virtual Hard Disk) Comparison Tool for Dynamic VHDs. Abhinav Aggarwal, Priyank Gaharwar, Vinod Kancharla)
Summer Intern: ARIES Lab, Department of Computer Science, University of Saskatchewan, Saskatoon, Canada, Summer 2011 (An Arduino Based Hardware Prototype for Audio Level Indication (Opencast Matterhorn Project). Abhinav Aggarwal, Jim Greer)
Teaching Assistance
CS 6432 – Distributed Consensus and Blockchains. (Instructor: Elaine Shi), Cornell University, Fall 2017.
EC 501 – Modeling and Simulation. (Instructor: Padam Kumar), Indian Institute of Technology, Spring 2014.
Talks
Price of Ignorance in Robotic Swarm Foraging
UNM CS Colloquium, Albuquerque, NM. Feb. 2019.
UNM STEM Research Symposium, Albuquerque, NM. Mar. 2019.
Scalable Sybil-Resistant Identity Generation.
WCEC (Workshop on Competitive Economics of Cybersecurity), Sandia National Labs, Albuquerque, NM. Nov. 2018.
On Some Lower Bounds for Oblivious Filtering on Data Streams.
UNM CS Colloquium, Albuquerque, NM. Aug. 2018.
Visa Research, Palo Alto, CA. Jul. 2018.
Agree to Encash! Blockchain-style Consensus From a Classical Viewpoint.
UNM Nano Café, Faculty Club, Albuquerque, NM. Mar. 2018.
UNM CS Colloquium, Albuquerque, NM. Feb. 2018.
Distributed Consensus and Blockchains: A Brief Tutorial.
(Invited talk) NetSec-2018, Indian Institute of Technology, Roorkee, India. Jan. 2018.
How to trust without any trust?.
UNM ECE Student Conference (NM-ECE-SC 2018), Albuquerque, NM. Apr. 2018.
UNM CS Student Conference (CSSC 2018), Albuquerque, NM. Apr. 2018.
CITP, Princeton University, NJ. Oct. 2017.
MIT-IBM J. Watson AI Lab, Cambridge, MA. Oct. 2017
Global SDN: Altruistic Attack Prevention.
Cornell University, Ithaca, NY. Nov. 2017.
University of Cincinnati, Cincinnati, OH. Oct. 2017.
MIT CSAIL, Cambridge, MA. Oct. 2017.
Visa Research, Palo Alto, CA. Aug. 2017.
Interactive Communication for Large Networks.
19th International Conference on Distributed Computing and Networking (ICDCN), Varanasi, India. Jan. 2018.
Visa Research, Palo Alto, CA. Jun. 2017.
UNM Stem Research Showcase, Albuquerque, NM. Feb. 2017.
Mid-South Theory Day, LSU, Baton Rouge, LA, Dec. 2016.
UNM CS Student Conference, Albuquerque, NM. Apr. 2016.
On the Equivalence of Probabilistic Automata.
UNM Term paper presentation, Albuquerque, NM. Apr. 2015.
Recursiveness, Computability and the Halting Problem.
Graduate Seminar, IIT Roorkee, India. Sept. 2013.