Editorial illustration for OpenAI's AI data agent, built by two engineers, now used daily by 4,000 staff
OpenAI's AI Data Agent Transforms Internal Workflow
OpenAI's AI data agent, built by two engineers, now used daily by 4,000 staff
Why does a tool built by just two engineers matter to a company of nearly 5,000 people? OpenAI’s internal AI data agent started as a modest experiment, a prototype meant to streamline how engineers retrieve and clean datasets. While the tech is impressive, its real impact shows up in adoption numbers that dwarf typical internal apps.
By the end of last quarter, more than 4,000 staff members were logging in each day, tapping the same interface to answer queries, generate reports, and even flag data quality issues. The rollout wasn’t a slow rollout; it was a rapid, company‑wide push that turned a niche utility into a daily habit for the majority of the workforce. That level of penetration is rare—especially for a home‑grown AI assistant.
Emma Tang, who heads OpenAI’s data infrastructure and led the two‑engineer team, explains how the project scaled so quickly and what it means for internal productivity. And it is now used by more than 4,000 of OpenAI's roughly 5,000 employees every day — making it one of the most aggressive deployments of an AI data agent inside any company, anywhere. In an exclusive interview with VentureBeat, Emma Tang, the head of data infrastructure at OpenAI whose team built.
And it is now used by more than 4,000 of OpenAI's roughly 5,000 employees every day -- making it one of the most aggressive deployments of an AI data agent inside any company, anywhere. In an exclusive interview with VentureBeat, Emma Tang, the head of data infrastructure at OpenAI whose team built the agent, offered a rare look inside the system -- how it works, how it fails, and what it signals about the future of enterprise data. The conversation, paired with the company's blog post announcing the tool, paints a picture of a company that turned its own AI on itself and discovered something that every enterprise will soon confront: the bottleneck to smarter organizations isn't better models.
Four thousand people rely on it daily. That's roughly 80 % of OpenAI's workforce. The agent emerged from a three‑month effort by two engineers, with AI writing seventy percent of the code.
When a finance analyst once spent hours sifting through seventy‑thousand datasets, a plain‑English Slack query now yields a chart in minutes. The company claims the approach is reproducible by any team, yet no external validation has been offered. Its rapid adoption makes it one of the most aggressive internal AI data agents reported, but whether similar speed and coverage can be achieved elsewhere remains unclear.
OpenAI’s data infrastructure lead, Emma Tang, described the rollout as a practical response to internal reporting bottlenecks rather than a polished product. The tool’s reliance on AI‑generated code raises questions about maintainability and oversight, especially as usage scales. For now, the agent functions as a daily workhorse for most staff, though its long‑term impact on workflow and data governance has not been measured.
Further Reading
Related Reading
Browse all OpenAI news →
Business & Startups OpenAI, a Series F San Francisco startup founded in 2015 by eight pioneers
Business & Startups Terminal-Bench 2.0 launches with Harbor, testing any container-installable agent
Business & Startups Zuckerberg Unveils Meta Compute to Build Global AI Infrastructure
Related Topic Gen AI app sessions up fivefold, downloads jump 778% as ChatGPT leads traffic
LLMs & Generative AI
Related Topic GPT-5 helps mathematicians offload tedious tasks, says Timothy Gowers
Open Source
From Archives Endor Labs launches free AURI tool after study finds only 10% of AI code is secure
From Archives Joe Gebbia seen with metallic device as OpenAI, Jony Ive partnership looms
From Archives Alibaba's Qwen3.5-9B outperforms OpenAI's gpt-oss-120B on laptop benchmarks
LLMs & Generative AI
From Archives OpenAI yields to Pentagon, bans bulk U.S. data; Amodei says law not yet
Common Questions Answered
How many OpenAI employees are using the AI data agent daily?
Over 4,000 of OpenAI's approximately 5,000 employees use the AI data agent every day. This represents roughly 80% of the company's workforce, making it an exceptionally high-adoption internal tool.
How was the OpenAI data agent initially developed?
The AI data agent was created by just two engineers in a three-month effort, with AI writing approximately 70% of the underlying code. The tool started as a modest prototype designed to help engineers retrieve and clean datasets more efficiently.
What practical impact does the AI data agent have on workflow efficiency?
The AI data agent dramatically reduces time-consuming data tasks, such as transforming a finance analyst's multi-hour dataset search into a quick Slack query that generates charts in minutes. This represents a significant improvement in productivity and data accessibility for OpenAI staff.