Back office ops · Production

Instacart builds Ava, a company-wide internal AI assistant powered by GPT-4

The problem

Instacart had no company-wide AI tool despite strong productivity gains during a hackathon with ChatGPT; non-engineering employees also faced a blank-text-box barrier that prevented broader AI adoption.

Workflow diagram · grounded in source
1
User submits request
trigger
“Users use Ava to write, to review and debug code, to improve comms, to learn faster and to build AI-enabled internal tools on top of our APIs”
2
GPT-4/3.5 generates response
ai_action
“an internal AI assistant powered by OpenAI's GPT-4 and GPT-3.5 models called Ava”
3
Slack thread summarization
ai_action
“we allow users to easily summarize a thread or channel by simply typing "@Ava summarize." The summary is then posted publicly”
4
Participants verify summary
human_review
“makes it easy for other participants to verify that the summary is accurate and note any corrections”
5
Prompt Exchange sharing
feedback_loop
“allow our users to create and share templates based on their own needs and experience”
Reported outcome

Ava reached company-wide adoption with over half of Instacart employees using it monthly and more than 900 weekly; the Slack thread summarization reduced a five-to-ten minute context grind to just a few seconds of skimming.

Reported metrics
Monthly active employeesover half of Instacart employees
Weekly active usersmore than 900
Average session duration20+ minutes per session
Hackathon feature output increasenearly twice as many features as initially planned
Show all 6 reported metrics
monthly active employeesover half of Instacart employees
weekly active usersmore than 900
average session duration20+ minutes per session
hackathon feature output increasenearly twice as many features as initially planned
Slack thread summarization time savedfive-to-ten minute context grind is reduced to just a few seconds of skimming
monthly adoption before Slack integrationnearly a third of the organization
Reported stack
AvaGPT-4GPT-3.5ChatGPTSlack
Source
https://tech.instacart.com/scaling-productivity-with-ava-instacarts-internal-ai-assistant-ed7f02558d84
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Ava reached company-wide adoption with over half of Instacart employees using it monthly and more than 900 weekly; the Slack thread summarization reduced a five-to-ten minute context grind to just a few seconds of ski…

What tools did this team use?

Ava, GPT-4, GPT-3.5, ChatGPT, Slack.

What results were reported?

Monthly active employees: over half of Instacart employees; Weekly active users: more than 900; Average session duration: 20+ minutes per session; Hackathon feature output increase: nearly twice as many features as initially planned (source-reported, not independently verified).

How is this back office ops AI workflow structured?

User submits request → GPT-4/3.5 generates response → Slack thread summarization → Participants verify summary → Prompt Exchange sharing.

Related back office ops cases