SERP Parser for SEO Analytics at SISTRIX

SISTRIX needed SEO analytics pipelines that could be built, shipped, and operated by the same team. A central part was the SERP Parser: collect search results across countries and devices, classify them, and make parser status visible instead of treating the system as a hidden batch pipeline.

I worked with a "You Build It, You Run It" model across Docker/Docker Swarm, CI/CD, Spark/Hadoop extraction for 450M+ keywords worldwide, and Apache Mesos/Marathon platform work. The parser made SERP Features such as organics, SEM, shopping, knowledge graph, maps, featured snippets, and sitemaps inspectable across countries and devices.

• SERP Parser: Made search results inspectable across countries, devices, and SERP Features
• Parser status: Play Framework dashboard for weekly runs, success states, throughput, and calc nodes
• Big Data Pipeline Development: HTML parser with XPath for millions of keywords per country, API integration
• Data Extraction: HTML crawler with Spark/Hadoop for 200M seeds, structured data extraction
• DevOps & CI/CD: "You Build It, You Run It" pipeline with Docker and automated deployments
• Platform setup: Built container setup, dashboards, and observability close to parser operations on provided server access
• PaaS Architecture: Apache Mesos/Marathon, AWS Route 53, scalable infrastructure
• Quality Assurance: Automated acceptance tests for SaaS tools with Cucumber and Selenium
• Monitoring & Operations: Status dashboard with Play Framework, operational transparency

• "You Build It, You Run It": End-to-end ownership for parser, pipelines, and operations
• Operational visibility: Parser status, SERP Features, throughput, and failure states turned runtime behavior into product feedback
• Big Data Processing: Spark/Hadoop and scalable data pipelines for repeatable search result analysis
• CI/CD: Automated tests, continuous deployment, and deployment feedback kept delivery tied to runtime behavior