Data Platform Case Study

Vista Analytics

Led the development of a centralized analytics platform serving internal teams, GTM functions, and customers through role-based access. Built data pipelines and dashboards that transformed leasing, operations, and AI assistant data into decision-ready insights. Instrumented AI leasing assistant and nurturing signals to support performance tracking and drive customer renewal conversations.

Company

Property Vista

Timeframe

Dec 2021 - Present

Snapshot

Hours → Minutes

QBR preparation time

Self-serve analytics

Customer access

Real-time visibility

Operational KPIs

AnalyticsData pipelinesRBACCustomer analyticsAI instrumentation

Overview

Vista Analytics transformed reporting into a product capability by turning fragmented operational data into reusable pipelines, governed data models, and self-serve analytics surfaces. The platform unified leasing, maintenance requests, purchase orders, insurance data, tenant demographics, and AI-driven leasing signals into a single analytics layer. It enabled internal teams and external customers to access consistent, decision-ready insights through embedded dashboards, Power BI integrations, and self-serve analytics tooling.

Problem

Operational and leasing data lived across disconnected systems, making reporting slow, inconsistent, and dependent on manual extraction and spreadsheet workflows.

Users / Stakeholders

Users

Executives and operators preparing QBRs and monitoring performance KPIs
Customer-facing teams requiring self-serve portfolio visibility
Product and operations teams making prioritization and optimization decisions

Stakeholders

Data and engineering teams (ETL, modeling, pipelines)
External customers using Power BI and embedded dashboards
Leadership requiring consistent cross-domain reporting

Constraints

Source data was fragmented and not originally designed for product-grade analytics
The solution had to support both internal reporting and customer-facing multi-tenant access
AI-assisted insights needed to be credible, explainable, and grounded in trusted data

My Role & Ownership

Defined the analytics roadmap across leasing, maintenance, financial, and AI data domains
Designed pipelines for maintenance requests, purchase orders, insurance, tenant demographics, and AI signals
Enabled multi-tenant access for internal and external users
Supported Power BI integrations and built internal analytics using Superset
Positioned AI insights as decision-support, not black-box automation

Approach

Started from business questions and KPI definitions before defining dashboards or outputs
Prioritized reusable data models and pipelines over one-off reporting requests
Designed analytics for adoption, trust, and governance as core product requirements

Solution

Delivered a centralized analytics platform built on reusable data pipelines, structured data models, and governed access controls. Unified leasing, maintenance, purchase orders, insurance, tenant demographics, and AI signals into one analytics layer. Enabled: - Superset dashboards for internal teams - Power BI integrations for customers - Real-time operational visibility

Technical Design

APIs, integrations, and data systems behind the product.

AWS Glue ETL pipelines for multi-domain data ingestion
S3-backed storage and analytics-ready datasets
Data modeling for tenant-aware, multi-tenant access
Power BI connectors for external analytics access
Superset dashboards for internal analytics
Pipeline orchestration for data freshness and trust

Tradeoffs

Prioritized data consistency and metric definition over rapidly shipping new dashboards
Delayed AI-facing features until underlying data reliability and trust were established
Balanced customer-facing flexibility with governance and multi-tenant access control

Metrics / Outcomes

Hours → Minutes

QBR preparation time

Replaced manual reporting workflows and spreadsheet stitching with structured dashboards.

Self-serve analytics

Customer access

Enabled stakeholders to answer core business questions without relying on custom data pulls.

Real-time visibility

Operational KPIs

Made leasing, funnel, and portfolio performance visible in near real-time.

Key Learnings

Analytics products fail when metric definitions are inconsistent, even if the UI is strong
Reusable data models create long-term leverage beyond individual dashboards
AI-driven insights are only valuable when grounded in trusted, operational data

External reference ↗