# How Plaid Built a Production SLA Dashboard in 3 Days with Replit Agent

Source: https://www.youtube.com/watch?v=VqZLg6ZhhOE
Recap page: https://rapidrecap.app/video/VqZLg6ZhhOE
Generated: 2025-11-11T19:34:57.423+00:00

---
## Quick Overview

Plaid successfully built a production-ready Service Level Agreement (SLA) dashboard in just three days during a 2-hour hackathon by leveraging Replit Agent to automate manual processes like uptime metric reviews, report generation, and data reconciliation, ultimately saving the team 96 hours annually and enabling a fully self-service model.

**Key Points:**
- A Plaid team built a production SLA dashboard in three days, stemming from a two-hour hackathon project.
- The automation project saved an estimated 96 hours per year (37 hours for SLA management + 20+ hours reconciling data/reports).
- The manual process involved reviewing uptime metrics via dashboard, generating reports, processing, and managing SLA processes, which was deemed quite manual.
- The solution, built using Replit, focused on automating uptime review, report generation, and data reconciliation, reducing manual time from 20 hours to 20 minutes.
- The resulting dashboard provides transparency, reduces friction, and enables a fully self-service model for internal teams (Customer Engineering, Support, Implementation, TAMs) and customers.
- The team decided against hiring consultants or buying generic software, choosing instead to have their internal team build the exact solution needed during the hackathon.

![Screenshot at 0:04: Text overlay announcing the project goal: "During a 2-hour hackathon" where the team decided to reimagine how they worked, leading to the automation effort.](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-00-04.png)

**Context:** Natalie Mercuri, Organizational Development Manager for the Customer Engineering and Support team at Plaid, discusses how her team sought to eliminate time-consuming, manual processes related to tracking Service Level Agreements (SLAs) and uptime metrics. The team identified that tasks like reconciling data and generating reports consumed significant time, prompting them to use a company hackathon to develop an automated solution using Replit Agent.

## Detailed Analysis

The presentation details how a team at Plaid, motivated by the tedious nature of managing Service Level Agreements (SLAs) and uptime metrics, used a two-hour hackathon to create an automated solution using Replit Agent. The existing manual process required over 20 hours monthly for reviewing uptime metrics, generating reports, processing data, and managing SLAs across teams like Customer Engineering, Support, Implementation, and Technical Account Management (TAMs). Initial estimates suggested this manual work consumed 37 hours/year on SLA management alone, plus over 20 hours/year reconciling data and reports, totaling a significant time sink. The team explicitly decided not to outsource this to consultants or use generic software, opting instead for an internal build. Using Replit, they developed a support dashboard that automated the review of uptime metrics, report generation, and data reconciliation. This automation slashed the time spent on these tasks from over 20 hours to just 20 minutes monthly, resulting in a saving of 96 hours annually. The new system drives transparency, reduces friction, and moves them toward a fully self-service model that customers are requesting, without requiring external intervention or waiting for IT.

### Problem Identification

- The manual process involved reviewing uptime metrics via dashboard
- Generating reports
- Processing and managing SLA processes, which was quite manual
- This consumed 20+ hours monthly for data reconciliation and report building.

### Hackathon Solution

- The team used Replit Agent during a 2-hour hackathon to automate the process
- They built exactly what they needed, avoiding consultants or generic software
- The goal was to create a visual representation of what customers need.

### Key Outcomes & Metrics

- Automation reduced manual time from 20 hours to 20 minutes for key tasks
- Total time savings estimated at 96 hours/year (37 hours for SLA management + 20+ hours for reconciliation)
- The dashboard promotes Transparency, reduces Friction, and enables Self-Service.

### Dashboard Features

- The Support Dashboard displays high-level metrics like SLA Compliance (78.21%), Avg Response Time (10h 5min), Total Tickets (25), and Uptime (99.67%)
- It features a Recent Activity log and a Credit Eligibility section.

![Screenshot at 0:00: Plaid logo displayed at the start of the presentation.](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-00-00.png)
![Screenshot at 0:04: Text overlay highlighting the context: "During a 2-hour hackathon" where the team decided to reimagine their work.](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-00-04.png)
![Screenshot at 0:05: Key metric highlight: The team aimed to save "96 hours/yr" by automating manual processes.](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-00-05.png)
![Screenshot at 0:12: The manual process steps are listed: "Review uptime metrics via dashboard" and "Generate reports".](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-00-12.png)
![Screenshot at 0:20: Visual confirmation that the existing process "is quite manual."](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-00-20.png)
![Screenshot at 0:55: Graphic illustrating the time sink: "20+ hours reconciling data & reports" with a loading bar.](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-00-55.png)
![Screenshot at 1:14: The result of the automation effort: tasks that used to take "20 hours" now take "20 minutes".](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-01-14.png)
![Screenshot at 1:34: The new Support Dashboard showing key metrics like SLA Compliance \(78.21%\) and Avg Response Time \(10h 5min\).](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-01-34.png)
![Screenshot at 2:47: Text emphasizing the proactive approach: "Plaid didn't wait for IT."](https://ss.rapidrecap.app/screens/VqZLg6ZhhOE/00-02-47.png)