Client Background
BA Property Tax is a Dallas-based firm serving real estate professionals, legal advisors, and tax consultants across Texas. Known for its expertise in navigating complex appraisal district systems, the firm supports clients by managing high-stakes property valuation, protest filing, and investment analysis. As the client base expanded, BA Property Tax needed a scalable digital solution to manage and normalize fragmented data from 15 county systems.
The Problem
Texas property tax workflows are complicated by a decentralized infrastructure. Each county’s Appraisal District operates a unique system with different interfaces, document formats, and search mechanisms. BA Property Tax teams spent hundreds of hours manually navigating county websites, extracting data, and verifying records. The process was slow, error-prone, and inefficient, limiting the firm’s ability to scale and creating operational bottlenecks during critical appraisal and protest periods. Existing automation tools lacked the adaptability to operate across bot-protected sites and diverse data formats.
Our Solution
NextGen Coding Company developed a purpose-built property tax automation platform tailored to BA Property Tax's precise operational needs. The solution spanned automation, intelligent data parsing, secure delivery, and robust infrastructure—resulting in a production-grade system capable of reliably scaling across Texas’s most complex appraisal districts.
Supported Counties
The platform automates data retrieval from 15 unique county systems, including Dallas, Tarrant, Bexar, Harris, Denton, and more. Each site required custom selectors, fallback logic, and format-specific handling for compatibility.
Automation Stack
Playwright (Python): Powers browser automation across JavaScript-heavy county sites.
Quart: Asynchronous backend API to manage jobs and streaming.
playwright_stealth: Browser stealth mode ensures access to bot-protected sites.
Tesseract OCR: Activated on-demand for image-based PDF parsing.
BytesIO Stream Processing: Enables in-memory handling of screenshots and PDFs.
asyncio + aiohttp: Support concurrent scraping jobs without blocking performance.
Automation Data Flow
The system begins with parcel list inputs. Using Playwright, the platform navigates CAD websites, extracts HTML and document data, and processes both through the Nanonets API, which returns structured JSON. A normalization module then unifies all output into a single schema. Final data delivery occurs via CSV, JSON, Google Sheets, dashboard, SFTP, or email.
Infrastructure Architecture
Amazon ECS / Fargate: Scalable microservices for compute.
Amazon API Gateway: Central entry point for backend communication.
Amazon S3 & CloudWatch: Storage and performance monitoring.
Admin UI: Web-based interface for administrators.
CI/CD Pipeline: Automated builds and deployment via Amazon ECR.
Results
Matched Data Extraction in 12 Counties: Full functionality confirmed during QA.
Upload & Process Success in All 15 Counties: Even in restricted environments.
Normalized JSON Schema: Eliminated manual cleanup and ensured instant usability.
Faster Processing: Screenshot streaming and real-time memory processing reduced latency.
Secure & Isolated Sessions: Maintained data integrity and avoided cross-county contamination.
The system now enables BA Property Tax to retrieve and process large-scale data with speed and confidence. It reduced manual workload, minimized errors, and improved turnaround times across county boundaries.
Call to Action
To streamline complex data retrieval systems or automate fragmented document workflows, Contact admin@nextgencodingcompany.com or book a call to speak with our solutions team to begin scoping
https://calendly.com/next_gen_coding_company/30min
Partner with a team experienced in deploying high-performance, fully customized automation across regulated sectors.
