1. Welcome to OUPI Data: Your AI-Native Data Hub

OUPI Data is the platform's AI-native data hub. Think of it as the central place where you import your tabular files — spreadsheets, CSVs — and then question, transform, and visualize them using plain language instead of code or SQL. It's also the foundation that other platform tools build on: Finance, Teams Intelligence, and even ChatPro all tap into datasets you store here. When you import a file, the AI automatically detects your columns and their types, so your data is ready to explore within seconds.

Once a dataset is imported, you can open the Query view and ask questions the way you'd ask a colleague: "total sales by region last quarter" or "top 10 customers by margin in 2025." Behind the scenes, the platform doesn't let a language model guess at the numbers. Instead, it translates your question into real code, runs that code on every row of your dataset inside a secure sandbox, writes a plain-language summary, and then verifies every number in that summary against the computed result. If anything doesn't match, you get a warning. The generated code, result table, summary, and any suggested chart are all visible and saved with the query.

Security is built into every query. Before any generated code executes, it's inspected: it cannot reach the network, access files outside your dataset, or touch the platform. Suspicious patterns are blocked and logged, and your original question is screened for manipulation attempts. The code then runs in an isolated sandbox. Your data never leaves the platform during this process.

Beyond querying, OUPI Data lets you transform data by describing what you need in plain language — filter rows, derive new columns, aggregate, or join datasets. The original dataset always stays intact; transformations produce refreshed views. You can also generate charts from any result and pin them to your dashboard to build a living overview. The Intelligence module watches your data continuously and raises alerts when it detects notable changes or anomalies, with a dedicated dashboard summarizing what moved and why it matters.

Regarding costs: importing and storing datasets is lightweight. AI-powered operations — querying, transformations, visualizations — consume credits per operation, and the cost is shown before you run. For very large datasets, the model samples the structure to understand your columns, but the actual computation always runs on all rows. If your analysis requires complex, multi-step modeling, consider using a mission where the AI can iterate over several steps.