Background

With over 2,500 students and $122 million in capital raised, BloomTech (formerly Lambda School) was one of Silicon Valley's hottest startups. To achieve this feat in under three years, Bloom leveraged an emerging tactic amongst startups: using no-code software. Specifically, the spreadsheet software Airtable and workflow automation software Zapier, were at the core of Bloom's early technology stack.

No-code tools like Airtable helped Bloom move fast at the start. However, once the student body grew to the thousands, they became more of a hindrance. Bloom needed help migrating from their no-code stack to more scalable technologies that could serve tens of thousands of students and beyond.

One of the highest leverage projects was building a tool to view any student's complete record.

*Internal tool only viewable by select Bloom employees with privileges.

Problem

Imagine clicking through a spreadsheet and it repeatedly freezing for up to a minute at a time. Now imagine happening to you dozens of times per day.

This was a daily frustration for employees. Most of the school’s information lived on one massive Airtable spreadsheet, containing hundreds of thousands of rows of data across 50+ tables.

The Student Operations team was one of the most frequent users of this spreadsheet. They looked up student records daily. Since the spreadsheet was so overloaded, and most of the student data existed across many tables, looking up one record took "between 5-15 minutes".

Solution

Student Operations needed a fast-loading custom tool to easily see a student’s complete record.

First, the engineering team had to build software to keep Airtable synced with a Postgres database. With the database ready, the next step was to build a custom tool on top of it.

Our team gathered requirements across the company regarding which datapoints were most important to surface. Then we constructed the tool’s interface using Retool, a rapid internal tool builder. We were able to from from conception to a shipped MVP within two weeks. Then we gathered feedback and refined the product -- including writing many large SQL queries, installing advanced security, and speeding up performance. We launched this large and complex feature company-wide in only another four weeks.

After orchestrating a handful of feedback cycles with end users to iterate on the tool, I recorded a video demo, wrote launch messaging, and announced the tool to the company as the Student Master Viewer ("SMV").

Impact

Saved 133 hours per week of manual productivity across 170 person team
Used over 100 times per day, as of April 2020

Quotes

Team

David Head - Team Lead
- Gathered requirements throughout the organization
- Built the MVP using Retool in two weeks
- Created a small beta group to test the tool before broad release
- Presented a live demo during All Hands and distributed a video demo

Brent Anderson - Backend Engineer
- Collaborated backend architecture to fit in with existing systems
- Collaborated on JWT security to integrate internal with existing auth system
- Collaborated on optimizing reads and writes using SQL and GraphQL
Josh Egan - Backend Engineer
- Collaborated backend architecture to fit in with existing systems
- Collaborated on JWT security to integrate internal with existing auth system
Matt Spector - Frontend Engineer
- Created internal-facing interface to embed SMV for the operations team to use
- Technical advising
Chandan Maruthi - Backend Engineer
- Query-building
- Technical advising
Himanshu Gahlot - Backend Engineer
- Wrote queries to
- Technical advising