Sai Yasaswini Majety - Senior Backend & Cloud Infrastructure Engineer

“As we celebrate our 35th year in operation, we couldn't be more grateful or proud of the endurance and resilience of our global Viasat team. To further show our appreciation, we are pleased to award you with an annual discretionary performance bonus award. As we continue to navigate these times, we are exceptionally grateful for the talented team we have helping us address technical and business challenges with creativity, leadership, and tenacity.”

Mark Dankberg (Executive Chairman) & Rick Baldridge (CEO)

Viasat Corporate Leadership: Commercial Aviation Resilience & Delta Fleet Delivery

Work Experience

Independent ResearchMar 2024 – Present

AI Systems & Cloud Infrastructure Researcher

Co-authored and published Indian Patent 202541026299 for modular deep learning architectures while engineering high-throughput RAG microservices and cloud infrastructure automation.

Python, FastAPI, Docker, Terraform, AWS ECS/EKS, LiteLLM, Vector Stores

View details

Viasat Inc.Jan 2023 – Feb 2024

Senior Software Engineer - I (Backend & Cloud Infrastructure)

Spearheaded Commercial Aviation ground-to-air microservices and real-time fleet telemetry, scaling platform availability to 99.99% across 1,000+ commercial aircraft (Delta Air Lines). Executed zero-downtime Multi-AZ RDS MySQL upgrades and automated enterprise TLS/SSL certificate lifecycles across 500+ edge gateways.

Python, FastAPI, AWS (ECS, RDS Aurora/MySQL), CloudFormation, Redis, MySQL, OAuth 2.0, Okta, Palo Alto, Aruba

View details

Viasat Inc.Jan 2021 – Dec 2022

Software Engineer II (Cloud Platform & API Microservices)

Accelerated multi-region cloud deployment velocity by 65% (from 4 hours to 45 minutes MTTR) by authoring modular AWS CloudFormation IaC templates and scripted Jenkins Groovy CI/CD pipelines. Elevated automated test coverage across billing and telemetry paths to >92% with Pytest.

Python, AWS CloudFormation, Jenkins, Groovy, Docker, Pytest, AWS RDS, Linux, Bash

View details

Viasat Inc.Jun 2018 – Dec 2020

Software Engineer I (Distributed Systems & Tooling Automation)

Built Commercial Air PAT (Pipeline & Telemetry Observability) dashboard in Flask and React, reducing release cycle verification latency by 40%. Standardized pre-merge Groovy regression pipelines across 50+ Jenkins nodes, and optimized CandEx 2.0 full-text search in Sphinx & Django (<180ms queries).

Python, Flask, React, Jenkins, Groovy, Robot Framework, MariaDB, Docker, Linux

View details

Viasat Inc.Jan 2018 – May 2018

Software Engineer Intern (Talent Systems & Search Automation)

Architected and deployed 'CandEx 2.0', an automated enterprise recruitment and candidate evaluation platform adopted across Viasat HR and technical hiring teams with Sphinx full-text search across 10,000+ candidate profiles.

Python, Django, Angular, MongoDB, Sphinx Search Engine, Apache Solr, MariaDB, REST APIs

View details

Enterprise & Research

Featured Projects

Production-grade distributed backends, aviation telemetry architectures, and deep learning patent implementations.

CandEx 2.0: Internal Developer Platform & Documentation Hub

Architected and owned the central enterprise developer portal and automated API catalog serving 500+ software engineers across Viasat. Built with Django, Angular, Sphinx, and PostgreSQL, containerized with Docker on AWS ECS. Reduced new-engineer onboarding ramp time by 65% and unified documentation across 60+ distributed services.

#django#angular#postgresql#sphinx#docker#aws-ecs

Modular Deep Learning Adapter Framework (Indian Patent 202541026299)

Engineered a novel parameter-efficient adapter framework enabling foundation models to transfer representations dynamically across disparate cross-domain distributions without catastrophic forgetting. Published with the Indian Patent Office (IPO). Features inter-layer routing adapters and catastrophic forgetting mitigation algorithms.

#pytorch#modular-dl#patent#transfer-learning#huggingface

In-Flight Telemetry Ingestion & Real-Time Fleet Pipeline

Designed high-throughput telemetry ingestion backend streaming connectivity health data from thousands of commercial aircraft in flight. Ingests 25M+ events daily with sub-second dashboard query latencies on AWS ECS, Kafka, and PostgreSQL. Supported 99.99% connectivity uptime SLAs for commercial airline carriers.

#python#fastapi#aws-ecs#kafka#postgresql#telemetry

Automated TLS/SSL Certificate Lifecycle Engine

Engineered automated certificate provisioning and rotation engine across 150+ microservice endpoints using HashiCorp Vault and Let's Encrypt / DigiCert APIs. Replaced manual multi-day renewal procedures with automated hot-reloading, eliminating certificate expiration outage risk completely.

#python#hashicorp-vault#tls-ssl#devops#automation

Enterprise Multi-Tenant OAuth 2.0 & Okta Identity Hub

Architected federated authentication and authorization gateway securing external airline client portals. Implemented dynamic token exchange, custom claims validation, and granular role-based access control (RBAC), satisfying strict aerospace enterprise cybersecurity compliance.

#oauth2.0#okta#security#jwt#rbac

IoT Smart Telemetry & Gateway Monitor

Hardware and cloud IoT monitoring solution capturing environmental and electrical distribution parameters via MQTT with real-time alerting. Won 1st Place in Viasat IoT Hackathon and served as the foundation for undergraduate research capstone at SASTRA University.

#iot#mqtt#python#microcontrollers#hackathon-winner

Patents & Deep Learning Research

Patent PublicationOFFICIALLY PUBLISHED (IPO)2025

Modular Deep Learning Architecture for Cross-Domain Transfer and Incremental Learning

Sai Likhith Kanuparthi, Sai Yasaswini Majety (Co-Inventors)

Indian Patent Office — Application No. 202541026299 · Government of India — The Patent Office Journal No. 13/2025

Architectures for modular, parameter-efficient deep learning adapters enabling foundation models to transfer representations dynamically across disparate cross-domain distributions without catastrophic forgetting. Formulates dynamic gating mechanisms, inter-layer routing adapters, and catastrophic forgetting mitigation algorithms. Classified under IPC: G06N 3/08 (Learning methods), G06N 3/045 (Value-driven neural networks), G06N 3/096 (Transfer learning), G06N 3/084 (Backpropagation), and G06N 20/00 (Machine learning).

Application No.202541026299

StatusPublished (The Patent Office Journal)

Filing DateMarch 21, 2025

Publication DateMarch 28, 2025

Publication IssueJournal No. 13/2025, Part 1

JurisdictionIndian Patent Office (IPO)

Verify on PatentAssist

Modular Deep Learning Architecture for Cross-Domain Transfer and Incremental Learning

Click to Enlarge

Honors & Key Recognitions

Indian Patent Published: Modular Deep Learning Architecture

Indian Patent Office (IPO 202541026299)

Parameter-Efficient Adapter Routing

March 2025

Verify Gazette Record

Dean's Merit Scholarship: Academic Distinction

SASTRA Deemed University

Top 3% Percentile (1,200+ Cohort)

2016-2017

1st Place Winner: Real-Time Fleet Telemetry Pipeline

Viasat IoT & Cloud Telemetry Hackathon

1st of 24 Engineering Teams

2019

Viasat Spotlight Award: Automated TLS/SSL Edge Security

Viasat Core Infrastructure Group

Zero-Outage SLA: 150+ Edge Gateways

2023

Architecture Commendation: Zero-Trust Identity Gateway

Viasat Architecture & Security Committee

Multi-Tenant Okta & OAuth 2.0 Federation

2022

Five-Nines SLA (99.999%): Zero-Downtime AWS Cloud Migration

Viasat Platform Operations

99.999% Telemetry Uptime: 45% Compute Cut

2022

Education & Academic Honors

SASTRAAug 2014 – May 2018

B.Tech in Electrical and Electronics Engineering (EEE)

Shanmugha Arts, Science, Technology & Research Academy (SASTRA Deemed University)

CGPA 8.2 / 10.0

Thanjavur, Tamil Nadu, India

4-year rigorous undergraduate engineering program spanning core electrical power systems, digital signal processing, microprocessors, embedded computing, and algorithmic problem-solving. Awarded the prestigious Dean's Merit Scholarship for ranking in the top 3% of the academic engineering cohort.

Dean's Merit ScholarshipDAKSH National Fest LeadIEEE Student Member

View Academic Highlights & Coursework

Contact & Opportunities

Get In Touch

LET'S CONNECT

I am actively seeking Senior Backend, Distributed Systems, and Cloud Infrastructure roles. Whether you are discussing engineering opportunities, satellite and telemetry architectures, or modular deep learning research - my inbox is always open.

Email Directly

yasaswini7777@gmail.com

Location & Availability

Houston, TX · Open to Relocation & Remote Roles

US Work Authorization

H-4 EAD (Authorized to work for any US employer without sponsorship)