> For the complete documentation index, see [llms.txt](https://amitagarwal.gitbook.io/profile/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://amitagarwal.gitbook.io/profile/research.md).

# Research Experience

### March 2025 - April 2025 : Research Institute on the Foundations of Computer Science (IRIF) - Paris Cité University, Paris

Topic: "**Pseudorandom Correlation Functions**" supervised by Geoffroy Couteau

* Design of efficient Pseudorandom Correlation Functions for practically useful correlations such as Oblivious Linear Evaluation (OLE) and Multiplication triples.

### May 2024 - August 2024 : Aptos Labs, USA

Topic: "**Fair dishonest majority secure computation using blockchain**" supervised by *Rex Fernando, Benny Pinkas, Alin Tomescu*

* Design of a new witness encryption scheme which enables efficient conversion of unfair dishonest majority MPC protocols into a fair protocol by leveraging blockchain in a lightweight way.

### May 2023 - August 2023 : Aarhus university, Denmark

Topic: "**Secure binary and arithmetic conversions**" supervised by *Prof. Peter Scholl*

* Design of new practically efficient schemes for binary and arithmetic conversions in secure computation and zero knowledge proof protocols.

### March 2023 - April 2023 : Technion - Israel Institute of Technology, Israel

Topic: "**Unit vector correlations and its applications**" supervised by *Prof. Yuval Ishai*

* New constructions of Pseudo-random Correlation Generators for boosting practical efficiency of secure computation tasks such as lookup table evaluation and private information retrieval.
* New protocols for secure sorting and selection via Function Secret Sharing.

### May 2022 - Aug 2022 : Google, USA - Research Intern

Topic: "**Integrating differential privacy and secure computation for large scale neural network training**" supervised by *Dr. Mariana Raykova and Karn Seth*

* Formalized new techniques for efficiently integrating central and local differential privacy inside secure two-party protocols for performing neural network training.

### May 2021 - Aug 2021 : Google, USA - Research Intern

Topic: "**Privacy Preserving Machine Learning**" supervised by *Dr. Mariana Raykova and Karn Seth*

* Design and analysis of new and efficient privacy-preserving machine learning algorithms using secure multiparty computation by leveraging novel cryptographic techniques.
* Implementation and benchmarking of the protocol using C++ in Google3 codebase

### Jun 2020 - Dec. 2020 :  Dept. of Computer Science, UIUC - Research Assistant

Topic: "**Two round maliciously secure computation"** supervised by *Dr. Dakshita Khurana*

* Deigned a novel maliciously secure MPC protocol which requires only two rounds of interaction between the parties, thus establishing a new state-of-the-art in the field of secure computation
* Wrote a full technical paper detailing the protocol and the security analysis. The work has been accepted at TCC 2021 and is [available on eprint](https://eprint.iacr.org/2021/1230.pdf)

### Aug. 2019 - May 2020 : Coordinated Science Laboratory, UIUC - Research Assistant

Topic: "**Robust and scalable schemes for Secure Multiparty Computation**" supervised by *Dr. Andrew Miller*

* Design and analysis of privacy-preserving secure multiparty protocols for applications involving multiple semi-trusted/untrusted parties e.g. Proof of Custody and Bank Clearance (Netting) problem.
* Implementation and benchmarking of designed protocol using [HoneyBadger MPC](https://github.com/initc3/HoneyBadgerMPC) and [MP-SPDZ](https://github.com/data61/MP-SPDZ) toolkit

### Dec. 2019 - May. 2020 :  Dept. of Computer Science, UIUC - Research Assistant

Topic: "**Post Quantum Secure Multiparty Computation in Constant Rounds"** supervised by *Dr. Dakshita Khurana*

* Explored non-rewinding based simulation techniques for proving security of protocols against quantum adversaries
* Design and analysis of a protocol in which multiple parties can generate unbiased and uniform randomness in a distributed fashion. The protocol is provably secure against a quantum adversary which can potentially corrupt all but one party. Additionally, the protocol only takes a constant number of rounds to complete.
* Wrote a full technical paper detailing the protocol and the security analysis. The work has been accepted at Eurocrypt 2021 and is [available on eprint](https://arxiv.org/abs/2005.12904)

### Aug. 2019 - Dec. 2019 :  Dept. of Computer Science,UIUC 

Topic: "**Fair protocols for Secure Multiparty Computation"** supervised by *Dr. Dakshita Khurana*

* Explored the characterization of non-binary and asymmetric functionalities which can (and cannot) be computed with fairness in settings without an honest majority.
* Construction of provably secure protocols (with fairness) for realizing such functionalities.

***

***

## Other research works done during my undergraduate

### Jan. 2018 - May 2018 : Nanyang Technological University, SINGAPORE - Research Intern

Topic: "**Secure virtual machine placement for IaaS clouds**" supervised by *Dr.Ta Nguyen Binh Duong*

* Designed a new virtual machine placement algorithm which combats side-channel attacks in public IaaS cloud data centers by reducing the probability of an adversary to co-locate their Virtual Machine(VM) with a benign user during initial VM placement. 
* Performed a theoretical and empirical analysis of the proposed algorithm to evaluate its resource efficiency and security. Results obtained using real-world Microsoft Azure workload dataset indicated that the proposed algorithm provides a significant increase in the cloud's co-location resistance with little compromise in resource utilization, compared to existing approaches.

### Aug. 2017 - Nov. 2017 : Birla Institute of Technology and Science Pilani, Goa, INDIA - Undergraduate student

Topic: "**Traffic Flow Confidentiality using TOR network**" supervised by *Dr. Sanjay K. Sahay*

* Studied the inner working of The Onion Routing (TOR) protocol and the strategy that it uses to reduce the risks associated with traffic analysis attacks.
* Investigated into some of the open research problems related to bridge (Tor relays that aren't listed in the main Tor directory) distribution inside the TOR network.

### May 2017 - Jul. 2017 : University of Southern California, Los Angeles, USA - Research Intern

Topic: "**Virtually Isolated DETER Experimentation**" supervised by *Dr. Clifford Neuman*

* Designed and implemented an infrastructure that enables remote isolated OS instances to connect with internal nodes for experiments in the DETER testbed (a cyber security testbed available to researchers worldwide).
* The project was used to extend the isolated domain of a DETER experiment to external virtual machines that are virtually isolated and capable of communicating only with the DETER testbed through an IPSec tunnel.

### Feb. 2017 - Apr. 2017 : Birla Institute of Technology and Science Pilani, Goa, INDIA - Undergraduate student

Topic: "**Clustering in ELM vs TF-IDF feature space"** supervised by *Dr. Rajendra Kumar Roul*

* Explored various concepts related to Extreme Learning Machines (ELM) and clustering algorithms.
* Performed a comparative analysis of standard clustering algorithms like K-means, Hierarchical Clustering and Fuzzy C-means clustering in ELM and TF-IDF feature space.

### May 2016 - Jul. 2016 : Indira Gandhi Centre for Atomic Research, Kalpakkam, INDIA - Research Intern

Topic: "**Hierarchical clustering algorithm for online resources**" supervised by *Mr. E. Soundararajan*

* Worked on hierarchical single linkage clustering using a 2 stage algorithm (a combination of Prim's and Kruskal's algorithm for minimum spanning trees). 
* Implemented the complete algorithm and tested it for 50,000 records of unordered books in SIRD Library, IGCAR.
* Designed a web-based interface for retrieval of clustered documents.