Lasantha Kularatne

AI/ML Engineering & Innovation

• Invented STAR (System Telemetry Analyzer), a GenAI-powered tool that collects metrics and uses LLM prompt engineering for root cause analysis, reducing Mean Time To Know (MTTK) by 80%
• Co-developed MCP Server enabling AI assistants to query live experiment data via natural language, reducing developer context-switching time by 80% and accelerating debugging 3x
• Built multi-agent LLM system using Strands SDK with OpenAI GPT for reasoning and Ollama local models for execution; presented to org-wide principal engineering community (blog post)
• Designed and executed LLM-based fraud detection experiments comparing XGBoost ML vs LLM performance on structured transaction data

Fraud & Risk Engineering (June 2025 - Present)

• Led Developer Productivity initiative achieving 100% Copilot adoption and 30-40% toil reduction through training, Champions group, and automation
• Guided fraud prevention teams (99.4% detection rate) on LLM-based detection techniques
• Automated vulnerability fixes using Renovate with AI adoption planned as next phase

Reliability Engineering (March 2020 - May 2025)

• Improved platform reliability from 98-99% to 99.9% (front-end) and 99.99% (back-end) for Tier-1 services through standardized SLI/SLOs and Error Budget Policies
• Built Clicker, a web analytics telemetry platform processing 100s of millions of Kafka events, achieving comprehensive SLI coverage for Shopping and Checkout
• Developed traffic prediction models using Fourier series achieving 95% accuracy for peak traffic forecasting across Expedia, Hotels.com, and Vrbo brands
• Built Traffic Watch system using AWS Lambda/S3/Datadog for long-term traffic data collection and analysis
• Created Expedia Group's first AWS Operational Readiness Review (ORR) template through workshops with AWS and architecture teams
• Designed and executed Rollback GameDay exercises for Tier-1 services, reducing change-related incidents by 80%
• Conducted SLO workshops across office locations, adapting Google's "Journey Through SLOs" training

Technical Leadership

• Received "Consistently Exceeds" rating with recognition from VP-level leadership as exemplary model for engineering engagement
• Mentored 10+ engineers to senior levels, providing career guidance and promotion preparation
• Founded cross-team reliability engineering community enabling knowledge sharing
• Conducted technical interviews at junior, senior, principal, and director levels
• Built relationships with Google, AWS, and Datadog through collaborative reliability initiatives

• Led design and development of Vrbo checkout microservice handling traveler transactions for web, mobile, and external point-of-sales during cloud migration
• Won first place in company-wide Kaggle ML competition for 'Booking Value Forecast' using Random Forest
• Sponsored technical interviews and mentored junior engineers through major platform transitions

• Led team building scalable web-app solutions for property owners and managers serving thousands of suppliers
• Designed scalable supplier analytics reporting API backed by Hadoop HBase, enabling real-time insights for property managers
• Created Solr-based fast search API for internal applications enabling sub-50ms lookups across millions of listed properties

• Built highly scalable REST API delivering ad snippets generated from user-generated content for online ads using NLP and machine learning
• Co-invented patented systems for dynamic insertion of user-generated content into advertisements (2 patents)
• Led legacy platform team and introduced Agile methodologies improving team productivity

Software Engineer II - Toppan Photomask | 2006-2009 | Round Rock, TX

Software Developer - University of Texas at Austin | 2006 | Austin, TX

Software Engineer - CodeGen.IT | 2003-2005 | Colombo, Sri Lanka