The person behind StoraIQ

I'm Txais Xiong.

Data analyst · Sacramento, California

I like the part of data work where you figure out why the records don't agree and make the next review easier.

My work has taken me from construction-permit reports to sewer assets and utility engineering records. The systems change. The responsibility stays familiar: understand the source, write down the rules and leave a process someone else can check.

StoraIQ brings that experience back to construction and housing. I'm starting here in Sacramento, building a place where you can explore public records, understand what a chart counts and follow the finding back to its source.

Explore the Sacramento dataConnect on LinkedIn ↗
Txais Xiong
Construction reporting is where I learned how much work happens before a number is ready to publish.

What CIRB taught me

Start with the work people actually do.

At CIRB, the Construction Industry Research Board program at the California Homebuilding Foundation, I began by entering permit data. Doing that work helped me see which decisions could become rules and which still needed a person. The pipeline I built followed that distinction.

  1. 01

    Rules for the clear cases

    Editable keyword rules handled familiar permit descriptions, so staff could change classification behavior without changing code.

  2. 02

    A model for ambiguity

    A PyTorch model used sentence embeddings (numerical representations of text) to help classify descriptions the rules did not settle.

  3. 03

    A person for the decision

    Low-confidence results went to a mandatory human review queue. The system gave uncertain cases a place to go.

Experience

The work behind the approach.

Feb 2026–Present

Pacific Gas and Electric Company

Engineering Assistant I · Data & Records Analyst

Contract through Beacon Hill Staffing Group

  • Audit and correct legacy engineering asset records in SAP dating back to 1980, 8,000+ completed to date, verifying construction milestone and project dates against engineering drawings during an active platform transition.
  • Apply business rules and structured quality checks that catch data-to-report mismatches before they reach supervisor-level reporting, where an incorrect milestone date would misstate project progress to leadership.
  • Standardize historical documentation against PG&E’s categorization schema so legacy records are retrievable and audit-ready; proposed a point-of-entry validation tool for drawing revisions and facility names.

May 2025–Jan 2026

Sacramento Area Sewer District

Data Management Technician

Contract

  • Owned the district’s weekly reconciliation between IT system exports and the GIS/AME (ESRI) asset database; built an automated Excel system that cut the audit cycle from 4 hours to under 15 minutes (95%, about 180 hours a year).
  • Replaced subjective manual review with a rule-based validation framework that auto-flags discrepancies before Quality Control, with an audit trail on every record.
  • Executed service-request change control in Maximo (CMMS), updating asset geometry, headers, indicators, and property lines to match engineering drawings; cross-referenced Granite inspection video against AME records to confirm pipeline lengths and cleanout locations.

Nov 2023–May 2025

California Homebuilding Foundation · CIRB

Senior Research Analyst

Data Entry Intern → Data Analyst → Research Analyst → Senior Research Analyst (Jan–May 2025)

  • Designed and shipped the statewide permit-classification pipeline: a JSON-configurable keyword rule engine for the clear majority of records, a PyTorch neural network on sentence embeddings for the ambiguous cases, and a mandatory human review queue for low-confidence results, covering 750,000+ permits from all 539 California jurisdictions in a 33-category taxonomy.
  • Cut processing time per report from 64.3 to about 20 minutes (about 69%), raised per-analyst throughput about 93%, and brought classification error rates from 20–25% to under 10%, replacing manual VBA/Excel workflows.
  • Built the system with self-taught Python and AI-assisted development, and externalized the rules so non-technical staff update classification behavior without code.
  • Adapted the pipeline to a new California Air Resources Board reporting agreement without added headcount; provided technical support and process documentation for staff and work-study analysts.

May 2021 - May 2024

Sacramento Chinese Community Service Center

Program Manager; previously After School Team Leader and Senior Team Leader

Program Manager: Oct 2023 - May 2024, concurrent with the CIRB internship. Team Leader roles: May 2021 - Oct 2023.

  • Ran a grant-funded after-school program for 100+ students, including staffing, scheduling, and compliance reporting.

Work history based on my September 2026 résumé.

Tools, with a purpose

What I bring to the work.

Records that agree

Data validation, business-rule checks, cross-system reconciliation and an audit trail for changes.

SAP · Maximo · ArcGIS / ESRI · Granite

Workflows people can run

Python and Excel automation, documented rules, clear handoffs and review queues for cases that need judgment.

Python · pandas · NumPy · Advanced Excel

Analysis people can inspect

Repeatable data preparation, classification and reporting, with the source and limitations kept close to the result.

Power BI · PyTorch · ETL workflows

Have a record that raises a question?

I'd like to hear what you're trying to understand about construction in Sacramento, or where the data leaves you stuck.

Get in touch