Open to IT management roles leading Agile delivery teams
Venkata Saranu
Principal Technical Lead · Engineering Leadership
I design and lead enterprise Java platforms — and the teams that build them. Twenty years across airline, telecom, healthcare and marketing data; the last ten at American Airlines, growing the Change Reservation journey and leading AI adoption.
20+ yrs engineering10 yrs at American Airlines12 engineers led
Ten years with American Airlines, rising from senior developer to Technical Lead to Principal Technical Lead. I architect microservices, coach teams through delivery transformation, and raise engineering standards with TDD, pair programming and CI/CD.
Today I lead about 12 engineers across two squads on the Change Reservation journey — and I'm looking for an IT management role leading Agile delivery teams.
Hiring, onboarding, feedback and mentoring engineers toward promotion
Architectural decisions for the Shopping Cart and Award domains
Adopting emerging tech — including AI skills the team uses every day
Celebrating 100 years of American Airlines
0+
years building enterprise Java platforms
0K→0K
Change Reservation transactions a day, with fewer rebook errors
0
engineers led across two onshore squads
0 mo
to deliver the net-new Award journey
The window shades lift as you scroll — tap one to close it.
Change Reservation volumetransactions per dayTeam shape12 engineers · 2 squads · XP practices
Career flight log
Twenty years, one through-line: complex systems that work
Career at a glance2004 – today
American Airlines · Fort Worth, TX
Principal Technical Lead
Technical Lead from Jan 2020 · Principal since May 2021
Lead ~12 engineers across two squads on the Change Reservation journey, growing volume from 7K to 20K transactions a day while reducing the rebook error rate.
Led AI research and implementation for the journey, and created proprietary AI skills for documentation and code maintenance that the team uses daily.
Delivered the net-new Award journey in 5 months; architected Award Fulfillment on Sabre Web Services with ITA/QPX flight search.
Reactive microservices on Spring WebFlux; OpenTelemetry and Fluent Bit into Azure Data Explorer, Dynatrace alerting, Splunk dashboards.
Coach the team in TDD, pair programming and Jenkins CI/CD; built a reusable Angular component library.
2020NOW
American Airlines · via Veritis Group
Senior Java Developer (Full Stack) & SQAD Lead
Jun 2016 – Jan 2020
Delivered the Choose Flights and Basic Economy modules of the online booking application.
Built gateway and Upsell microservices (Spring Boot, Redis, Keycloak, Hystrix) so customers can upgrade from Basic Economy to Main Cabin.
Jenkins pipelines to IBM Cloud; Unaccompanied Minor proof of concept in Mendix with Sabre and an Android QR scanner.
20162020
AT&T · Plano, TX
Senior Application Developer
May 2014 – Jun 2016
Built the DME2 / GRM / LRM service infrastructure with load balancing and failover for web, JMS and JDBC services; modeled routing data in Cassandra.
20142016
Epsilon · Irving, TX
Java / Big Data Application Developer
Jun 2013 – May 2014
Helped design the Customer Data Store on Cloudera Hadoop (HBase, Hive, Impala) loading client files with millions of records; proofs of concept for Storm, Spark and Cascading.
20132014
Earlier flights · 2004 – 2013
Enterprise Java across four industries
Blue Cross Blue Shield of Texas — enterprise quoting and enrollment (Spring MVC, Apache CXF, JRules).
XO Communications — service-activation systems provisioning orders on network devices.
Verizon Business — IMPACT workflow platform automating network-alarm handling.
Sun Microsystems India — marketing and trademark portals (Struts, EJB, Hibernate).
20042013
Cruising altitude
Leadership first, then the toolbox
PRE-FLIGHT CHECKLISTCAPTAIN · LEADERSHIP
Leading and mentoring Agile teams — Agile / Scrum / XP, sprint planning
Mission-critical delivery and delivery transformation
Hiring and onboarding; coaching engineers toward promotion
Emerging-technology and AI adoption
Architectural decision-making; requirements and JAD sessions
TDD and pair programming; business demos
ALL ITEMS CHECKED · CLEARED FOR TAKEOFF
PROFESSIONAL LICENCE
Holder
VENKATA SARANU
Education
B.Tech, Computer Science & Information Technology Jawaharlal Nehru Technological University · 2004
Ratings
Sun Certified Java Programmer Sun Certified Web Component Developer (Java EE 5)
A1ARCHITECTUREMicroservicesSpring BootWebFluxSpring CloudREST & SOAPApigee EdgeSabre Web ServicesCircuit breakersON TIME
B2CLOUD & DEVOPSIBM Cloud KubernetesCloud FoundryAWSJenkins CI/CDOpenTelemetryFluent BitAzure Data ExplorerDynatraceSplunkBOARDING
C3LANGUAGES & DATAJava / J2EETypeScriptAngularOracleMongoDBRedisCassandraHadoopON TIME
D4QUALITYJUnitMockitoCucumber BDDJMeterTDDCLEARED
Featured project · American Airlines
AA Reservation Assistant — an AI agent that changes and cancels trips
A conversational agent built on LangGraph's ReAct pattern. Customers change or cancel a reservation by chatting on the web or by messaging from their phone; the agent reasons about each request, calls the right tools, and iterates until the job is done. I led the architecture, design and implementation.
Figures from the project brief for the multi-step form flow the agent replaces.
How the agent thinks — and a live walk-through
Every message is scrubbed of personal data, then the LLM reasons, picks a tool, observes the result and loops. Run a conversation and watch a plane fly each hand-off through the graph.
Web chat · Reservation Assistant (demo)
ReservationsText message · mobile
Interactive illustration with fictional booking data — same conversation mirrored on web chat and mobile messaging.
Tools the agent can call
Identity verificationCollect PNR, name and date of birth; 3-attempt lockout.
Reservation lookupFetch the full booking and show a summary.
Flight searchFind alternatives, check changeability, support multiple dates.
ML rankingScore and filter the options before the customer sees them.
Select & pricePick a flight and show the fare breakdown.
Confirm & bookTake payment and confirm the change with new tickets.
CancelApply fare rules and cancel with the refund or credit.
Clarify / switchAsk follow-ups or switch to a different reservation.
One agent brain, every channel
💬
Web chat
Embedded as an iframe with a single script tag, or run standalone. Rich flight cards and inline payment.
📱
Mobile app
WebView integration with push notifications and the same streaming experience.
✉️
Mobile messaging
WhatsApp and SMS through gateway adapters: same tools, same identity checks, same protections.
PII never reaches the LLM
Customer says“Change PNR ABC123, last name Smith”
LLM sees“Change PNR [REDACTED], last name [REDACTED]”
Real values live in a server-side vault. The model only sees which fields were collected — tools read the vault directly. Verification locks after three attempts.
Server-Sent Events let customers watch the agent search and compose in real time.
Ranking the options — from a week of data
Hundreds of alternatives can come back for one trip. A learned ranking model scores each on price difference, schedule shift, stops, seat availability and cabin match; hard rules filter out unusable ones first. It works from a rolling 7-day window of availability and booking outcomes.
