# What vBase Verifies

vBase creates **globally credible, independently verifiable, proof** of what data existed — and when.

Each **stamp** serves as a cryptographic commitment linking a specific digital object (usually a file) to a point in time and to the dataset, or **Collection**, it belongs to.

When combined into a **Collection**, these stamps establish an independently verifiable record of the creation and revision of the dataset — turning historical datasets into auditable assets.

## Core Dimensions of Verification

vBase verifies three fundamental aspects of data history and presentation:

### **1. Timestamp Integrity**

vBase proves the exact moments when data, files, or results within a dataset or strategy were created or updated.

Each stamp includes a **cryptographically verifiable timestamp**, recorded immutably on a public blockchain. This allows any third party to confirm *when* specific data became known — eliminating ambiguity about version histories or backdating.

**Example:**

* A data vendor can prove that their June 2022 data has not been revised since June 30, 2022.
* An investment analyst can prove when they published a winning stock call and the corresponding investment model
* A risk officer can prove which risk report the company saved at the end of each month of the past year.

### **2. Dataset Integrity**

vBase ensures that the files or data within a Collection have remained **unaltered since their original timestamps**.\
By comparing the hash of the current data to the blockchain-backed audit trail for the Collection, anyone can confirm that:

* No extra data has been inserted.
* No data has been removed.
* The data is *exactly* as it was when first stamped.

This gives investors, partners, or clients cryptographic assurance that historical data matches exactly what it was in the past — not modified, selectively cleaned, or incompletely queried.

**Example:**

* A research firm can demonstrate that they are sharing every research report they wrote about Microsoft over a given historical period
* An investment manager can demonstrate that they are sharing all historical trade files in a strategy, with no extra or missing trades
* A scientist can demonstrate that all lab results recorded at the time are being presented for testing and evaluation to journal referees

### **3. Impartial Presentation**

Beyond individual files, vBase verifies that a dataset or trading strategy is not being selectively presented from many parallel alternatives.

By creating a public record of the number of datasets a given user maintains, vBase makes it clear **what population the presented result came from**.

This addresses the classic “data dredging” or “cherry-picking” problem — proving that what’s shown is representative of the research universe, not the lucky survivor.

**Example:**

* A quant researcher can show that their live strategy came from one of three tested models — not one of hundreds — giving allocators confidence that performance wasn’t selectively chosen.
* An experimental data provider can show he didn't run multiple versions of the same experiment, cherry-picking the most suitable result
* An investment analyst can show he didn't publish multiple versions of his investment model, and then selectively presented the one closest to the final outcome

***

## What This Enables

Together, these verification dimensions create a **globally credible, independently verifiable audit trail** for any dataset, model, or portfolio.\
This audit trail shows *exactly what data existed, and when* — making data history verifiable and trustworthy.

With vBase, data producers and consumers can:

* **Prove data timeliness** — demonstrate that claims or results were known at the time stated.
* **Prove data integrity** — verify that data and results are original and unmodified.
* **Prove impartiality of presentation** — show that findings weren’t cherry-picked from a larger unseen set.

The result is transparent, verifiable data provenance — turning “trust me” into **“verify it yourself.”**