Sachin Araballi

• Design, develop, and maintain scalable backend services and distributed data pipelines supporting Target.com search and fulfillment systems serving millions of daily users.
• Led end-to-end engineering for Get It Today / Get It Tomorrow and Circle 360 offer filters using REST APIs, Apache Solr indexing, and Kafka-based data ingestion pipelines.
• Developed Agentic Search capabilities using LLMs and LangChain technology to extend natural-language search and relevance workflows for e-commerce discovery.
• Led technical strategy for Solr Priority Indexing during peak traffic events, implementing a gold/silver indexing model to surface mission-critical data first — presented at InnerCon 2025.
• Led design for personalized Dynamic Promo Landing Page ranking using Member ID and Visitor ID signals to improve guest-level relevance.
• Created new indexing logic using multi-channel attributes and qualifier-based signals, contributing to a 0.4% relative uplift in search conversion.
• Co-led GenAI Innovation Week, including MOSS++, a hybrid evaluation platform for Recommenders and Search using LLM-based diagnostics; earned 3rd place in the Relevancy Hackathon.
• Enabled MLOps adoption with Vertex AI, including GPU provisioning, CI/CD integration patterns, and production-readiness practices for scalable model deployment.
• Built internal GitHub Metrics Extractor tooling to surface engineering delivery insights for leadership.
• Mentored engineers through design reviews, onboarding, and production workflow guidance.

• Designed and implemented secure payment integration services using Node.js, contributing to measurable growth in credit-card payment revenue.
• Built API Gateway architecture and ELK-based observability pipelines to improve API security and monitoring.
• Developed Python-based data processing solutions to detect missing marine terminal gate activities from EDI data.

• Led development of a Data Matching Tool using Apache Spark, Helidon microservices, and Oracle Cloud Infrastructure to compare more than one million customer records at scale.
• Conducted applied ML research using SVM-based de-duplication models to identify duplicate customer records during onboarding.
• Designed and implemented a Data Profiling Platform using Elasticsearch and PL/SQL to analyze more than 20 million records and improve enterprise data quality.
• Performed advanced Solr performance tuning in cloud environments and mentored campus hires.

• Designed and implemented Java-based plugins for Tizen Studio supporting OSGi and Eclipse-based Smart TV and IoT application development.
• Built C# language support using OmniSharp and the Language Server Protocol to improve developer tooling.
• Developed IoT device certification workflows for Samsung Artik 530 to enable secure integration with SmartThings Cloud.
• Served as Scrum Master, leading agile ceremonies and sprint planning.

• Integrated Apache Solr search services with Oracle ADF-based user interfaces to improve enterprise search responsiveness.
• Designed reusable Oracle Fuse-based UI components to maintain architectural consistency across enterprise products.

Production-grade ML pipeline for entity resolution using a Support Vector Machine with 15 pairwise similarity features. Achieved F1 = 1.0 on the FEBRL4 benchmark and F1 = 0.85 on adversarial test cases. Served via a FastAPI inference endpoint with structured JSONL logging.

Cross-platform fitness app with category-filtered workout library, a finite-state-machine exercise/rest timer, and animated progress tracking. Built with a composite navigation pattern (Bottom Tab + Native Stack) and declarative Animated API.

Full-stack multi-tiered parking management system demonstrating SOLID principles and design patterns (Strategy, Factory, Repository). Backend supports multi-floor configuration, vehicle-type routing, dynamic pricing, and persistent JSON storage. Comprehensive REST API with a React dashboard.