miranda murarik↗
← All projects
Project collection

Data engineering

Recurring research and reporting need data that holds up to scrutiny. My work includes the pipelines that move it and the checks that catch problems along the way.

CONFIGCONFIGCONFIGCONFIGCONFIGCONFIGCONFIGCONFIG
VALIDATE
shared engine
8research practices
Data engineering / Analytics

Maintaining a shared validation engine

The research practices had nearly identical validation scripts that were getting hard to maintain. I replaced them with a shared engine. Adding a practice now means updating its configuration.

The engine surfaces roughly 200 data-quality flags per full refresh cycle, based on my own estimate.

How I approached it +

I kept member data out of the AI coding workflow. The model saw only structure and aggregate diagnostics, and I used synthetic data to check the code. Comparing the results with a colleague’s manual count uncovered a join bug.

Claude Code · Python · Data validation · Synthetic fixtures
How the workflow connects
Inputs
  1. FilesBenchmark CSV exports

    Tall and wide response tables

  2. ConfigPractice configuration

    Validation rules and instrument settings

Processing
  1. CodePython · pandas

    Check outliers, zeros, and missing values

Outputs
  1. FilesExcel workbooks

    Flagged values and review decisions by practice

Analysts decide whether to keep, remove, or correct each flagged value. This validation step leaves source files unchanged.

APIAlteryxBigQuery
Quarterlyrefresh ownership
Data engineering / Analytics

Owning a quarterly financial-data pipeline

I took over a recurring financial-index refresh and moved it from manual CalcBench entry into an API-driven Alteryx pipeline feeding BigQuery.

The work included the refresh itself, data-quality checks, documentation, and outlier detection in Tableau.

How I approached it +

The pipeline still needed someone to check the data each quarter. I owned that process alongside the migration, so the automation and the checks were maintained together.

Claude Code · Alteryx · BigQuery · Tableau · APIs
How the workflow connects
Inputs
  1. APIFinancial Modeling Prep API

    Financial statement data

Processing
  1. WorkflowAlteryx

    Transform data for the recurring index refresh

Outputs
  1. WarehouseBigQuery

    Financial index data tables

  2. ReportingTableau

    Dashboards and outlier review from the warehouse

The pipeline owner reviews data quality and outliers during the quarterly refresh.

Data access / Research tools

A read-only research instrument browser

Part of the benchmark platform

I built a tool for browsing published benchmark instruments, filtering questions, and exporting modules to Excel.

How I approached it +

I used a read-only API so a colleague could explore the instrument without the tool being able to change the source content.

Claude Code · Python · Contentful · Excel
How the workflow connects
Inputs
  1. Read-only APIContentful Delivery API

    Published instruments, modules, and questions

  2. InputAnalyst request

    Practice, question type, or module selection

Processing
  1. CodePython

    Traverse instruments and filter questions

Outputs
  1. FileExcel

    Selected instrument content, organized by module

The client is read-only: browsing and export cannot change the published content.

Data updates / Content operations

Updating benchmark instruments from spreadsheets

Part of the benchmark platform

I built a write-enabled tool that creates or updates question, practice, and metric definitions in the CMS from Excel, with a result log for each row.

Claude Code · Python · Excel · Contentful
How the workflow connects
Inputs
  1. FileExcel

    Question, practice, and metric definitions

Processing
  1. CodePython

    Read rows and create update payloads

Outputs
  1. Write APIContentful API

    Created or updated instrument definitions

  2. FileRow-level log

    Success or failure for each input row

This is a write-enabled tool. The result log records which rows succeeded or failed.

Reporting / Survey tools

Poll reports and survey dashboards

Part of the poll lifecycle

I created skills for preparing poll results reports and survey dashboards as part of the survey workflow.

Claude Code · Reusable skills · Survey reporting
How the workflow connects
Inputs
  1. InputCompleted poll results

    Data from the closed poll

Processing
  1. AI instructionsClaude Code skills

    Guide report and dashboard preparation

Outputs
  1. OutputPoll reports

    Prepared results for reporting

  2. OutputSurvey dashboards

    Prepared dashboard material

These reusable skills support the reporting stage of the survey lifecycle.