# Aayush Soni

> Software Development Engineer II @ Amazon

Location: New York, New York, United States
Profile: https://flows.cv/aayushsoni

## Work Experience
### Software Development Engineer || @ Amazon
Jan 2024 – Present | New York, NY
Helping improve conversational shopping experience, product discovery, and Advertising using "Sponsored Prompt" Ads and Rufus LLM - a Gen AI powered Shopping Assistant.

### Software Development Engineer II @ Amazon Web Services (AWS)
Jan 2022 – Jan 2024 | Seattle, Washington, United States
AWS Contact Lens team.

### Software Development Engineer I @ Amazon Web Services (AWS)
Jan 2021 – Jan 2022 | Seattle, Washington, United States
AWS VoiceID team.

### SDE Intern, AWS AI @ Amazon Web Services (AWS)
Jan 2020 – Jan 2020
Interned in the AWS Contact Lens team. Contact Lens is a service providing NLP-powered analytics for call-center conversations, helping businesses improve their customer support services.

- Developed a new ML service, capable of performing real-time inference on text data using an NLP model, as part of an upcoming feature of AWS Contact Lens. Technologies used include Python, PyTorch, Docker and AWS Sagemaker, among others.
- Automated service deployments by writing infrastructure code to create a production-ready CD pipeline along with the AWS infrastrcuture needed to support the service.
- Surpassed the internship goals by partly completing the integration of this service with existing system, in addition to predetermined deliverables.

### Graduate Research Assistant @ Rochester Institute of Technology
Jan 2020 – Jan 2020 | Rochester, New York Area
- Developed probabilistic models (Bayesian HMMs and Coupled Bayesian HMMs) for modelling time-series feature sequences of converstions between two persons.

### Graduate Research Assistant @ Rochester Institute of Technology
Jan 2019 – Jan 2020 | Rochester, New York
- Worked on a Deep Learning research project aimed at generating high-quality, realistic aerial imagery from low-quality, cheap and easy-to-produce synthetic aerial imagery (image-to-image translation).
- Performed image pre-processing and wrote code for testing various state-of-the-art Neural Network (Conditional GAN) architectures on Multispectral Image data.
- Identified Pix2PixHD as the best baseline and migrated its PyTorch code to the latest Tensorflow 2.0 to speed up model training/inference time by taking advantage of compiled graphs.
- The paper describing this work has been published in a special session on "Drone Imaging" in "Imaging and Multimedia Analytics in a Web and Mobile World Conference '20" held at San Francisco, CA.


## Education
### Master of Science - MS in Computer Science (Intelligent Systems Track)
Rochester Institute of Technology

### Bachelor of Engineering - BE in Information Technology
Gujarat Technological University (GTU)


## Contact & Social
- LinkedIn: https://linkedin.com/in/aayush1196
- Website: https://github.com/aayushARM

---
Source: https://flows.cv/aayushsoni
JSON Resume: https://flows.cv/aayushsoni/resume.json
Last updated: 2026-04-05