> ## Documentation Index > Fetch the complete documentation index at: https://docs.onebucket.io/llms.txt > Use this file to discover all available pages before exploring further. # Introduction > One S3- and MCP-native endpoint to all your data — across cloud, on-prem, and edge, with no copying or rearchitecting. OneBucket gives your AI and applications a single endpoint to **all** your data — across cloud, on-prem, and edge — over the two protocols they already speak: **S3** and **MCP**. No copying, no migrations, no rearchitecting. ## Why OneBucket Most organizations can put less than 10% of their data in front of an AI model. The rest is scattered across clouds, on-prem systems, and the edge — too costly to move and too slow to use, so projects stall at the proof-of-concept stage. OneBucket presents every backend as one logical bucket, so your data stays where it lives and becomes usable everywhere. ## How it works OneBucket is built on the **Adaptive Data Fabric** — location-independent access to unstructured data: * **Connect natively** — native S3 and MCP interfaces reach every cloud, on-prem, and edge source, with no application changes. * **Place by policy** — policy-driven placement and adaptive movement keep data in the right place at the right time. * **Access anywhere** — intelligent caching and replication make distributed data feel local. ## What you can do * **Use it as S3** — point any S3 client or SDK at your OneBucket endpoint. Full AWS SigV4 auth, buckets, objects, multipart uploads, versioning, and presigned URLs. * **Give AI native access** — connect Claude or any MCP client to browse, read, and manage objects and query storage activity in natural language. * **Span every backend** — federate AWS S3, Azure Blob, Google Cloud Storage, Wasabi, Akave, Snowflake, Databricks, on-prem, edge, and air-gapped stores behind one namespace, then copy, migrate, or locate objects across them. * **Cut egress and latency** — cache at the edge and prefetch hot objects so reads are local. * **Govern and audit** — enforce one policy-based access control and stream detailed storage events. ## Get started Connect an S3 client or Claude and read and write your first object in minutes. Give Claude native access to your buckets over MCP.