Senior Software Engineer at Atlassian (2025-07 – Present)
- Contributed to full-stack enterprise search and AI-assistant features for Jira and Confluence using React, TypeScript, Java/Spring Boot, Python services, OpenSearch, Redis, Docker, and Kubernetes, supporting retrieval across millions of indexed content items.
- Developed permission-aware search and source-reference workflows using REST and GraphQL APIs, BM25, vector retrieval, metadata filters, and authorization checks, improving top-result relevance by about 12% in offline evaluations.
- Optimized backend orchestration and conversational APIs through Redis caching, request batching, asynchronous processing, and query tuning, reducing p95 response latency by about 18% while maintaining auditability and tenant isolation.
- Partnered with product, security, platform, and machine-learning engineers through design reviews, code reviews, testing, and production retrospectives, helping reduce post-release defects by about 15% and improve delivery predictability.
Senior Full-Stack AI Engineer at Practice AI (2022-10 – 2025-06)
- Delivered multi-tenant legal-medical web applications and document workflows using React, TypeScript, Python/FastAPI, PostgreSQL, Redis, Celery, Azure services, OCR, search, and LLM integrations, reducing first-pass case-preparation effort by more than 50%.
- Built secure case and document-management features with reusable React components, REST APIs, RBAC, encrypted Azure Blob Storage, audit trails, background processing, and status dashboards, reducing case setup and document organization by about 30%.
- Developed a document-grounded demand-letter workflow with upload, review, editing, source references, prompt versioning, validation, and attorney approval, reducing typical drafting time from 3-5 hours to about 30-45 minutes.
- Implemented a scalable case-summary and chronology pipeline using Azure AI Document Intelligence, FastAPI, PostgreSQL, pgvector, Redis, Celery, extraction rules, confidence thresholds, and reviewer flags, cutting first-pass review time by about 60%.
- Integrated external case-management systems through OAuth 2.0, REST APIs, webhooks, Azure Service Bus, idempotency controls, retry queues, and reconciliation jobs, achieving more than 99% synchronization success after automated retries.
- Collaborated with attorneys, paralegals, product managers, and client-success teams to refine interfaces, validation rules, and processing workflows from production feedback, reducing recurring extraction corrections by about 20% and onboarding time by 25%.
Software Developer at Optum Global Solutions (Philippines) (2018-03 – 2022-08)
Developed full-stack healthcare applications for claims review, anomaly detection, clinical document processing, and care management using Angular, React, Java/Spring Boot, Python, PostgreSQL, Oracle, Redis, and containerized services across a four-year progression.
- Built Angular dashboards and Spring Boot APIs for claim queues, payment holds, notes, approvals, and supporting-document requests, reducing average review time by about 22% across payment-integrity workflows.
- Implemented validation rules, eligibility and provider integrations, scheduled ETL jobs, Redis caching, field masking, and audit logging, automating initial validation for about 55% of claims and reducing incomplete submissions by 15%.
- Developed Python data pipelines and inference services for claims anomaly detection using XGBoost, random forests, logistic regression, isolation forests, Spark, and Airflow, scoring about 1-2 million claims per month.
- Created React investigator dashboards and explainable APIs with Flask, Spring Boot, PostgreSQL, Redis, contributing factors, and reviewer feedback, reducing low-value manual reviews by about 25% without automating final claim decisions.
- Built a clinical document-review workflow using React, FastAPI, ClinicalBERT, BioBERT, spaCy, Celery, and human validation, achieving about 85% precision for common entities and improving coder throughput by 20%.
- Delivered member risk and care-gap features using React, Spring Boot, FastAPI, XGBoost, SHAP, Kafka, PostgreSQL, and Redis, helping teams prioritize 10,000-20,000 members monthly and reducing dashboard response time by 35%.
Junior Software Engineer at ByteDance (2017-10 – 2018-01)
- Supported internal operations and content-management tools using Python, JavaScript, SQL, APIs, and web frameworks, delivering automation scripts that reduced repetitive reporting and data-cleanup work by about 8 hours per week.
- Resolved API, frontend, testing, and data-quality issues under senior-engineer guidance, completing more than 20 production fixes while building foundations in code review, monitoring, and release practices.
