# Codex Powers Data Science with Interactive Reports _OpenAI's Codex is shown to act as a data analytics agent, querying Databricks and generating interactive reports to diagnose business issues like booking cancellations._ **Published:** 2026-06-09 **Source:** https://www.startuphub.ai/ai-news/artificial-intelligence/2026/codex-powers-data-science-with-interactive-reports --- The latest advancements in AI are rapidly transforming how we interact with data. A recent demonstration showcases how Codex, OpenAI's advanced AI model, can be leveraged as a powerful tool for data science. By acting as an agent, Codex can directly query data sources like Databricks and then generate interactive reports, significantly streamlining the analysis process. Data Science ChallengesDriver manual effort in data analysis and report generationFrom the article 3 mentionsA recent demonstration showcases how Codex, OpenAI's advanced AI model, can be leveraged as a powerful tool for data science.OpenAI CodexCoreFrom the article 9+ mentionsA recent demonstration showcases how Codex, OpenAI's advanced AI model, can be leveraged as a powerful tool for data science.acts asData Analytics AgentContextFrom the article 4 mentionsBy acting as an agent, Codex can directly query data sources like Databricks and then generate interactive reports, significantly streamlining the analysis process.generatesInteractive ReportsEffectgenerates visualizations for business issue diagnosisFrom the article 9+ mentionsThe goal is to create an interactive report that diagnoses a recent spike in booking cancellations for a hypothetical company called Wanderbricks and recommends next steps.Streamlined WorkflowOutcomereduces manual effort for data scientistsFrom the article 2 mentionsThe video illustrates a practical application of Codex within the data science workflow.Booking CancellationsContextexample business issue diagnosed with reportsFrom the article 5 mentionsThe report pinpoints short lead-time bookings as the strongest indicator, with a 32.8% cancellation rate for bookings made less than 7 days in advance.leads toActionable InsightsOutcometransforms raw data into digestible reportsFrom the article 3 mentionsThis capability moves beyond simple data retrieval, enabling users to create sophisticated visualizations and derive actionable insights with far less manual effort. This capability moves beyond simple data retrieval, enabling users to create sophisticated visualizations and derive actionable insights with far less manual effort. The demonstration highlights how Codex can transform raw data into easily digestible and manipulable reports, saving valuable time for data scientists and analysts. The full discussion can be found on **OpenAI Youtube**'s YouTube channel. ![](https://img.youtube.com/vi/Lvk_VZOppIY/maxresdefault.jpg) Codex for data science, from OpenAI Youtube ## Codex as a Data Analytics Agent The video illustrates a practical application of Codex within the data science workflow. The system is prompted to use a data analytics plugin to query Databricks data. The goal is to create an interactive report that diagnoses a recent spike in booking cancellations for a hypothetical company called Wanderbricks and recommends next steps. Codex then executes a series of tasks, including querying Databricks directly, exploring files, running commands, and binding charts to SQL files. This process, which would typically take data scientists hours, is accomplished by Codex in a matter of minutes. The AI agent is shown to handle data refreshes, identify outliers, and perform trend analysis, culminating in a comprehensive report that includes executive summaries, detailed charts, and actionable recommendations. ## Interactive Reporting and Data Exploration A key feature demonstrated is Codex's ability to generate interactive reports. Users can explore charts, view data sources, and even switch chart types within the generated report. This interactivity allows for deeper dives into the data and provides greater flexibility in how the information is presented and analyzed. The example report focuses on Wanderbricks' cancellation rates. It reveals a significant spike in July, with cancellation rates rising from 20.4% in June to 28.8% in July. The report pinpoints short lead-time bookings as the strongest indicator, with a 32.8% cancellation rate for bookings made less than 7 days in advance. Codex also provides recommendations, such as validating the window before changing operations and monitoring daily cancellation rates. The system's ability to display the source query and preview data rows further enhances transparency and trust in the generated insights. This level of detail allows users to understand how the report was created and to potentially replicate or adapt the analysis themselves. ## Streamlining the Data Science Workflow The ultimate value proposition shown is how Codex can significantly accelerate the data science workflow. By automating the tasks of data querying, analysis, visualization, and report generation, Codex frees up data professionals to focus on higher-level strategic thinking and decision-making. The demonstration concludes with Codex's capability to export the generated report into various formats, including Google Slides. This feature allows for easy sharing and presentation of findings to leadership, further integrating AI into business processes. The tool aims to make data analytics more accessible and efficient for a wider range of users. --- Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory.