Srinivas Kalyan YellankiRESEARCH & ENGINEERING

SRINIVAS KALYAN YELLANKI · AUSTIN, TEXAS

Srinivas Kalyan
Yellanki

I’m a software engineer and applied researcher working on the reliability of distributed systems and AI personalization infrastructure.

Selected publicationsGet in touch
11+ years in engineeringIEEE Senior MemberM.S., UNLV

PROFESSIONAL MEMBERSHIPS & RESEARCH SERVICE

Selected recognition
IEEESenior Member2025 — present ASSOCIATION FOR COMPUTING MACHINERYProfessional Member2025 — present SIGMA XIFull Membership2026 — present
2026 SERVICEICSEArtifact EvaluationASEArtifact EvaluationISSREArtifact EvaluationICPC NACJudge

SELECTED PRODUCTION IMPACT

Prior work at Expedia Group
7.5B+SDK requests per month
across the serving platform
$290KEstimated annual
infrastructure savings
25%Reported improvement
in developer productivity
30%Reported reduction
in onboarding time

Figures are reported from my prior work; financial and productivity outcomes are estimates.

01 / SELECTED RESEARCH

From production questions
to testable evidence.

My work connects operational experience with empirical methods and reproducible studies.

2026AISQ · ISSREPresented
First author
with Nicholas Piazza

AI PERSONALIZATION · SOFTWARE RELIABILITY

Operational Reliability Patterns in AI Personalization Serving Infrastructure: Practitioner-Grounded Hypotheses and a Clean-Room Reproduction Study

Turns five practitioner observations into measurable, refutable hypotheses. An accompanying clean-room benchmark reproduces tail latency amplification caused by a missing SDK outer timeout boundary.

Reproducibility artifact Workshop
Contribution & study scope

My contribution. Led the research framing, formulated the hypotheses, designed and conducted the reproduction study, analyzed the results, and prepared the manuscript and public artifact.

Study scope. The benchmark examines one generalized serving mechanism. It is a reproduction study, not a direct empirical validation of all five hypotheses, and uses no proprietary production data.

Venue context. AISQ 2026 is a specialized, peer-reviewed workshop on advanced intelligent software applications, co-located with ISSRE. Held on October 20, 2026 in Limassol, Cyprus, it provided a focused setting for discussing AI software quality and reliability. This is a workshop publication.

Peer-reviewed full workshop paper · Presented October 20, 2026 · Limassol, Cyprus · AISQ 2026, co-located with IEEE ISSRE 2026 · DOI activation pending

2014ICPCPublished
Co-author
Empirical software engineering

HUMAN FACTORS · PROGRAMMING LANGUAGE DESIGN

What Is the Foundation of Evidence of Human Factors Decisions in Language Design? An Empirical Study on Programming Language Workshops

A systematic examination of how programming language research uses empirical evidence about people, based on a corpus of 359 papers from programming language workshops.

Andreas Stefik, Stefan Hanenberg, Mark McKenney, Anneliese Andrews, Srinivas Kalyan Yellanki, and Susanna Siebert.

Paper & DOI
Contribution & publication details

My contribution. Participated in the systematic review and manual coding of the paper corpus while working with Professor Andreas Stefik at UNLV, including research review sessions and classification work.

Publication. Proceedings of the 22nd International Conference on Program Comprehension, pp. 223–231. The study acknowledges NSF grant CNS-0940521.

Venue context. ICPC is an established research conference devoted to understanding software through methods, tools, and empirical studies. Its 22nd edition took place June 2–3, 2014 in Hyderabad, India, co-located with ICSE 2014. This research conference is distinct from the ICPC programming competition.

ICPC 2014 conference details ↗

MASTER’S THESIS · UNLV · 2015

A Survey on Potential Privacy Leaks of GPS Information in Android Applications

Examines location privacy risks in Android applications and methods for detecting and preventing information leakage.

Read thesis

02 / ENGINEERING EXPERIENCE

Reliability, in practice.

More than a decade building the services, data paths, and developer tools that production systems depend on.

Expedia Group

Software Engineer 3 · Personalization

JAN 2023 — APR 2026

Advanced the Personalization SDK from versions 2 through 4 and developed context services for recent search history and lodging filter intent. Standardized typed JVM, non-JVM, and GraphQL integration contracts and failure handling across consumer experiences, helping teams adopt and contribute to the shared SDK. Developed the Personalization Context Service and home / search APIs connecting traveler signals to serving decisions.

  • Supported a serving platform handling more than 7.5 billion SDK requests per month.
  • Built Kafka ingestion into Cassandra / Scylla and bounded context reads through time-window partitions and server-side aggregation. Optimized Spark / Airflow pipelines, including a loyalty-context job reported to fall from 6–7 hours to under 30 minutes.
  • Built Chrome-based introspection and early conversational debugging tools using retrieval and tool invocation. Added Datadog tracing, SLO burn alerts, and Kubernetes autoscaling; mentored engineers and an intern in the Austin pod.
SDK designContext infrastructureKubernetes
Selected outcomes

Reported outcomes include approximately 25% higher developer productivity, 30% faster onboarding, 30% less troubleshooting time, an estimated $290,000 in annual infrastructure savings, and improvements associated with over $2 million in annual incremental bookings. These reflect contributions within a broader team and platform.

Cisco

Software Engineer · Webex Contact Center Data Platform

AUG 2021 — JAN 2023

Built the ingestion and transformation layer behind Webex Contact Center’s multichannel communication analytics. The data platform processed approximately 20 million events per day, turning operational events into information for analytics and reporting.

  • Developed Java / Spring Boot microservices and Kafka / Flink pipelines for real-time ingestion and transformation.
  • Improved ETL throughput and end-to-end processing latency across the data path.
  • Added dashboards and distributed logging to investigate incidents, identify processing failures, and support platform stability.

Focus: timely, observable data processing for contact center operations.

Stream processingData platformsProduction diagnostics

John Deere Financial

Software Engineer · Financial servicing

AUG 2020 — AUG 2021

Developed customer and servicing workflows for credit, installment loans, and equipment leasing, connecting a React application to Java / Spring Boot microservices.

  • Supported application modernization on AWS using EC2, SQS, and Lambda.
  • Contributed to data migration while preserving consistency and limiting disruption to financial servicing workflows.
  • Improved stability and performance across the client, API, and service layers.

Focus: dependable financial applications through cloud and data transitions.

Cloud modernizationAWSApplication reliability

Walmart

Software engineering · Workforce management

JUL 2015 — JUL 2020

Developed workforce planning, labor allocation, and performance analytics systems for distribution centers. These systems connected millions of daily operational events with staffing and decision support.

Modernized application capabilities through Java / Spring services and messaging with Kafka, IBM MQ, and ActiveMQ, while coordinating delivery with contract engineers.

  • Applied event-time ordering and idempotent recency controls so delayed or duplicate events could not overwrite newer decision state.
  • Built replay and reconciliation capabilities for broker outages and reprocessing, with freshness monitoring and decision provenance to support recovery and auditability.
  • Developed a handheld application for real-time workforce tracking on MC40 / MC70 devices.
Event-time correctnessReplay & reconciliationDecision provenance

University of Nevada, Las Vegas

Graduate research & teaching

AUG 2013 — MAY 2015

Conducted empirical software engineering research with Professor Andreas Stefik and supported C++ instruction as a graduate teaching assistant.

  • Contributed to the systematic review and manual coding of a 359-paper corpus for the published ICPC 2014 study on human factors evidence in language design.
  • Participated in research review and classification sessions connecting programming language decisions with empirical evidence.
  • Completed a master’s thesis examining potential GPS privacy leaks in Android applications.

Focus: empirical research, programming education, and mobile privacy.

03 / OPEN SOURCE

Work you can inspect.

Public artifacts make the assumptions, methods, and implementation available for others to examine.

RESEARCH ARTIFACT

Personalization Serving
Reliability Benchmark

A clean-room companion to the AISQ 2026 paper. Compares timeout configurations to examine how an outer serving boundary affects caller-visible tail latency.

run.shinner_only.csvouter.csvsummary.txt
Public reproducibility artifactView repository
ONGOING DEVELOPMENT

IntentTrace

An early-stage Kotlin / JVM project exploring execution budgets, context correctness, and decision provenance. The goal is to make serving behavior measurable across retries, fallbacks, and dependency boundaries.

BudgetContextProvenance
Kotlin / JVM · In developmentView repository
THE NEXT QUESTIONS

How should execution budgets compose? When is context too stale to use? What evidence explains a decision after the fact?

04 / SERVICE & RECOGNITION

Contributing beyond
the implementation.

Research artifact evaluation, competition judging, technical presentations, and professional community participation.

2026

ICSE

Artifact Evaluation · Program Committee

Committee

48th IEEE/ACM International Conference on Software Engineering

Rio de Janeiro, Brazil · April 12–18, 2026

A premier international conference for software engineering research and practice, spanning the methods, tools, and evidence used to build and maintain software.

My role: Artifact evaluation—examining research artifacts for functionality, reproducibility, and consistency with reported results.

Conference & venue details ↗
2026

ASE

Artifact Evaluation Committee

Committee

41st IEEE/ACM International Conference on Automated Software Engineering

Munich, Germany · October 12–16, 2026

A premier research forum for automated software engineering, covering automation in software analysis, design, implementation, testing, and maintenance.

My role: Artifact evaluation—examining research artifacts for functionality, reproducibility, and consistency with reported results.

Conference & venue details ↗
2026

ISSRE

Artifact Evaluation Committee

Committee

37th IEEE International Symposium on Software Reliability Engineering

Limassol, Cyprus · October 20–23, 2026

An established international forum specializing in software reliability and dependability, bringing together research and industry perspectives on trustworthy software systems.

My role: Artifact evaluation—examining research artifacts for functionality, reproducibility, and consistency with reported results.

Conference & venue details ↗
2026

ICPC North America Championship

Judge · Collegiate programming competition

Competition

SELECTED TECHNICAL PRESENTATIONS / 2025

DATA & PERSONALIZATION

Generative AI in Action: Building Personalized Travel Experiences from Clickstream Signals

Beyond the Cloud: Innovations in Big Data & Engineering Intelligence

AI & DEVELOPER TOOLS

Conversational AI Meets Enterprise Search

Workshop on Agentic AI & AI Innovations · University of the West Indies / Quantum Insights

FOUNDATIONS

Education

2015

M.S., Computer Science & Engineering

University of Nevada, Las Vegas

2013

B.Tech., Computer Science & Engineering

Jawaharlal Nehru Technological University

LET’S CONNECT

Good systems start
with good questions.

For research discussions, technical collaboration, or a closer look at my engineering work, I’d be glad to connect.

y.srinivaskalyan@gmail.com
LinkedIn GitHub

Citation

Select and copy the citation below.