Hi, I’m Darshil 👋
I'm a software engineer at Amazon Web Services (AWS). I graduated with a Master's degree from Carnegie Mellon University where I was advised by Professor Ranysha Ware and Professor Srinivasan Seshan. My academic/research interests lie at the intersection of congestion control and low-latency streaming. Beyond that, I like to work on projects in the field of networking, distributed systems, software engineering, and compiler design.
Outside of work, I enjoy going on long drives. I also like to cook Indian food (which almost always ends up not looking like the photo on the recipe website).
Research
VCCAnalyzer: Identifying Congestion Control Algorithms used by Video Streaming Platforms 🔗
Carnegie Mellon University, Pittsburgh, PA
January 2024 - May 2025
Work Experience
Software Engineer
Amazon Web Services, Seattle, WA
July 2025 - Present
- Part of EC2 Nitro team
- Handling packet processing in Elastic Network Adapter (ENA)
Teaching Assistant - Networking and the Internet
Carnegie Mellon University, Pittsburgh, PA
August 2024 - December 2024
- Teaching (and learning) networking concepts!!
- Concepts include internet architecture at scale, congestion control, BGP, datacenter networks, and much more… 🔗
Software Engineer
Exclaimer, Bloomington, IN
March 2022 - November 2022
- Created a scalable sandbox deployment system for dev, staging, and prod environments
- Mentoring interns on the scheduling team
- Working on the scheduling algorithm
Software Engineer
Periodic, Bloomington, IN
June 2021 - March 2022
- Working on the core scheduling team and payment processing team.
- Led the development of an MVP during the acquisition of Periodic by Exclaimer.
Education
MS in Information Networking
Carnegie Mellon University, Pittsburgh, PA
August 2023 - May 2025
BS in Computer Science
Purdue University, West Lafayette, IN
January 2018 - May 2021
Minor in ECE
Projects
Kernel Introspection with Wasm 🔗
- Integrated a wasm runtime into the Linux kernel to allow eBPF-like packet processing functionality
- Preliminary testing shows promising performance, closely approaching the efficiency of eBPF
- Future work includes optimizations such as support for JIT compilation and zero-copy networking
P4CKET: P4 Program Optimization through Online Profiling 🔗
- Introduced an algorithm to adaptively instrument P4 program based on the real-time traffic profile
- The compiler is modified to include passes which augment the program adding instrumentation and modifying the match-action logic
- Evaluation indicates reduction in stage memory allocation by up to 40%
TCP Reno Implementation in C
- Implemented the TCP Reno congestion control algorithm in C from scratch
- Features include fast retransmission, congestion avoidance, and fast recovery mechanisms
Other Projects
TAC0: A Compiler for C0 in Rust 🔗
- Designed a compiler for C0, a stripped down version of C
- Supports optimizations like tail call elimination, function inlining, branch prediction, and SSA-based optimizations
- Cross-compilation to RISC-V also available
CCAnalyzer in Mininet 🔗
- Implemented the CCAnalyzer paper in Mininet to support running experiments in parallel
Raft Consensus in Go
- Implemented the Raft consensus algorithm in Go, ensuring fault tolerance and consensus in networked systems
Mixnet Routing in C
- Designed a privacy-preserving mix network overlay in C to anonymize web traffic and enable secure communication over the Internet, incorporating encryption, redirection, and mixing techniques to safeguard user privacy
Get In Touch!
LinkedIn
GitHub✉️ Email - Username = dkaneria, Domain = cmu.edu