RUCHI SAHA

Software Engineer

India5 yrs 4 mos experience

AI ML PractitionerAI Enabled

Key Highlights

Filed a patent for a novel KAG algorithm.
Stabilized a service handling over 3 billion requests per day.
Winner of Microsoft’s Global Hackathon.

Stackforce AI infers this person is a Backend Engineer with expertise in high-throughput systems and cloud services.

Contact

Skills

Core Skills

Microsoft AzureLarge Language Models (llm)React.jsRedisMicroservicesAwsData ModelingFpga

Other Skills

Semantic SearchKnowledge Graph AugmentationTypeScriptContent Distribution NetworksC#Representational State Transfer (REST)Azure DevOps(ADO) PipleineScalaProgrammingAmazon Web Services (AWS)GitKubernetesAgile MethodologiesRelational DatabasesPostgreSQL

About

Backend engineer and system‑builder obsessed with scale, performance, and innovation. With 5+ years of experience, I’ve built and scaled high‑throughput systems, including services handling 7+ billion requests per day and platforms processing 100GB+ of data daily. Beyond scale, I’m deeply driven by innovation and research—from filing a patent, to winning Microsoft’s Global Hackathon, to contributing through technical research submissions. I thrive at the intersection of engineering excellence, experimentation, and real‑world impact. Complementing my backend focus, I also bring hands‑on frontend experience from working on the UI of Microsoft’s flagship products, enabling me to build systems that are not only scalable, but also product‑aware and user‑centric.

Experience

5 yrs 4 mos

Total Experience

3 yrs 3 mos

Average Tenure

2 yrs 1 mo

Current Experience

Microsoft

Software Engineer II

Apr 2024 – Present · 2 yrs 1 mo · Noida, Uttar Pradesh, India

Filed a patent on a novel KAG (Knowledge Augment Generation) algorithm for enabling deep domain specific or cross-domain research in a private knowledge base.
Redesigned and rebuilt critical web UI components for Microsoft powerhouse applications (Word, Excel, PowerPoint, OneNote), introducing new components and modernized architectures that achieved a 60% performance improvement in component-level boot and render time.
Worked on Word, Excel, and PowerPoint home pages, contributing to the design and implementation of server-side rendering architecture, and led security and privacy reviews to ensure compliance and robustness.
Led simultaneous cross-team service takeover and onboarding for two projects from different teams with a 2-person core team; drove knowledge transfer, authored documentation, managed customer escalations, and coordinated access/setup during a high-ambiguity platform migration. Served as primary on-call for two months to ensure operational stability during transition.
Post-handover, onboarded 10+ new team members to the service and implemented standardized on-call, incident management, and deployment processes, improving operational readiness and reducing production risk.
Stabilized a 3B+ requests/day high availability service by reducing failure alerts by ~97% (900 to 30 per day) and failed HTTP requests (HTTP 5xx) by ~99% (8.6M to less than 98K per min) during peak traffic. This service provides user specific license and entitlements data to 110+ Microsoft web apps (e.g. Word, Excel, PPT, Outlook, Sharepoint etc)
Innovation - Winner, 3rd Place, in Microsoft’s Global Hackathon.

Microsoft AzureLarge Language Models (LLM)Semantic SearchKnowledge Graph AugmentationReact.jsTypeScript+5

Samsung r&d institute india

Engineer, Backend Server

Jan 2021 – Apr 2024 · 3 yrs 3 mos · Noida, Uttar Pradesh, India · On-site

Designed and developed microservice based softwares, which collects data from several continents to empower Samsung TV+ services. Added features for improved monitoring of data quality.
Developed a new system to process data in any json format for expansion of core business. Also ensured easy integration of the newly created system into existing ones.
Redesigned core systems which improved data correctness to ∼100% and also provided a better scalability with reduced processing time. The system processes more than 100Gb of data every day.
Fixed legacy issues in a critical service which minimized 5xx errors on servers. The system uses CDN and serves more than a million requests a day directly to active Samsung’s Smart TVs.
Performed failure identification and recovery at pod level in Kubernetes (infrastructure level using AWS CLI).
Innovation: Awarded ’Best In Development - Business Idea’ at the organizational idea-demo. Won a prize for a chosen concept in the department-level ideathon. Conducted multiple events to promote innovation culture in the department. Focused on inter and intra team innovations as well as innovations for Samsung TV+ services.

ScalaData ModelingRedisProgrammingAmazon Web Services (AWS)Microservices+13

Indian institute of information technology design & manufacturing kancheepuram

2 roles

Graduate Research And Teaching Assistant

Oct 2019 – Jun 2020 · 8 mos

Tech Stack:Python, FPGA (Programmable hardware), Verilog, Vivado, Network System Design, IPv4 Packet Forwarding
Developed a FPGA-based Multi-match Packet Classification Architecture for Network Intrusion Detection Systems.
Achieved 5.9x higher performance while consuming 5x lower resource. Also enabled configuration time hardware scaling.
Work submitted to the journal of Computer Systems Science and Engineering.
Supervised by Dr Noor Mahammad SK, supported by Ministry of Electronics and Information Technology, Govt. of India.

PythonFPGAVerilogVivadoNetwork System DesignIPv4 Packet Forwarding

Research Intern

Jul 2019 – Oct 2019 · 3 mos

Designed framework, assembly language, assembler and simulator (for assembly) for a customized hardware.

RoutingResearchProblem SolvingField-Programmable Gate Arrays (FPGA)LinuxProgramming Concepts+1

Institute for development and research in banking technology (idrbt)

Research Intern

May 2019 – Jul 2019 · 2 mos · Greater Hyderabad Area

Tech Stack: Python, Cryptography
Researched on Data Integrity and Deduplication in fog to cloud based IoTs, Supervised by Dr P Syam Kumar.
Work was submitted to the Computer Networks journal in Elsevie

ResearchProblem Solving