Knowing how your tenant’s Delivery API is used in practice helps you get an overview of what's being requested and tune your clients. Which views are requested most? Is an old link still being called? Which frontend sends the most calls? 

The new Delivery API usage page in the Management App answers these questions. You'll find it under Settings → Delivery API usage.

Examples of what you can see

• Requests per day – total Delivery API traffic over the last 7 or 30 days
• Most-requested URLs and handles – the content your applications fetch most often
• Not-found requests – URLs that were requested but didn't match a route

You can split each of these by environment and environment client.

Why it matters

Splitting by environment client shows which application is sending the requests. Your production website, your mobile app and a test script no longer appear as one combined total. Some examples:

• Spot a spike in not-found requests after a migration → find the URLs → fix the routes or add redirects
• Check if a handle or URL is being called more than required from your frontend
• See whether a development environment client is sending more traffic than you expected

What you need to do

Nothing. The page is available to all tenants now, with no setup required.

Questions? Reach out on your support Slack channel or at support@enterspeed.com.

Connecting Enterspeed MCP servers to your AI agent just got way easier.🔌

A new Connect an agent page in the Management App (under Settings) gives you one-click setup or ready-to-copy setup instructions for connecting the Management MCP or Query MCP to the agent you use.

Pick your agent, copy, done

Click Connect an agent and choose Management MCP or Query MCP. Then select your agent from the list:

• Agent CLIs – Claude Code, Codex, Gemini CLI, Copilot CLI
• Web – Claude.ai, ChatGPT
• Desktop apps – Claude Desktop
• IDEs – Cursor, VS Code

Each agent gets instructions in the format it expects. For Claude Code, that's a single terminal command:

claude mcp add enterspeed-management --transport http "https://mcp.management.enterspeed.com/"

For the Management MCP, you can also turn on Read-only before you copy. With it on, the connection refuses every write tool, which is useful when you want an agent to look but not touch. The setting is part of the copied instructions, so set it before copying.

One connection for all your tenants

The connection signs in as you rather than as a specific tenant, so a single connection gives you access to every tenant you're a member of. You pick which tenant to work with during the conversation and can switch whenever you like. Access matches what you have in the Management App. 

A quick recap of the two MCP servers

• Management MCP lets an agent read and change your tenant configuration: sources, mapping schemas, deployments, destinations and more. Ask for a tenant overview, investigate processing errors, or have the agent draft and dry-run a mapping schema.
• Query MCP lets an agent search and read the content in your indexes, answering questions grounded in your real data. The first time you use it in an environment, it creates an environment client there called Enterspeed MCP (sign-in) and reuses it from then on. Index tools show up once a tenant and environment are selected, so ask the agent to pick one first.

Getting started

Simply sing into the Enterspeed Management App, open Settings → Connect an agent and follow the instructions for your agent.

Read more in the Enterspeed MCP documentation.

Most tenants run more than one environment side by side – development, staging, production. From now on, they all keep making progress, even when one of them is busy.

For a long time, Enterspeed has enforced fairness between tenants, so no single tenant can monopolise processing and hold up everyone else on the platform. We've now brought that same principle between environments within a single tenant.

Previously, jobs ran through a single first-in, first-out queue for the whole tenant. So a large backlog in one environment sat in front of everything else, and jobs from other environments had to wait their turn behind it. Now Enterspeed picks jobs from every environment that has work waiting. If development queues a large batch – a bulk import, a full re-processing, a schema experiment – production jobs added afterwards no longer sit behind that entire backlog. Each environment keeps flowing.

Why it matters

Your environments now stay responsive independently of one another. Development can queue as much work as it likes without pushing production to the back of the line – production keeps processing alongside it. Environments in the same tenant still share one pool of processing capacity, so a genuinely busy tenant can still mean things run a little slower overall, but one environment can no longer block another.

What you need to do

Nothing. Environment-level fairness is already live and applies automatically to every tenant. There's no configuration to enable and no changes needed to your ingest, schemas, or environments.

Questions? Reach out on your support Slack channel or at support@enterspeed.com.

Want's to ingest data to Enterspeed – products from your PIM, orders from your ERP, content from your CMS. Now you can wire that data into automated workflows without writing a single line of glue code. ⚡

We're excited to announce the official, verified Enterspeed node for n8n – the workflow automation platform that connects 1000+ apps and services. With the Enterspeed node, you can push data into Enterspeed and pull it back out as part of any automation, all from n8n's visual canvas.

It's available now, so go ahead and give it a try.

Why it matters

n8n is where teams stitch their tools together – triggering flows on a schedule, reacting to webhooks, and moving data between systems. Adding Enterspeed to that mix means you can automate the unglamorous parts of your data supply chain: sync a product feed on a timer, ingest CMS changes the moment they happen, or fan data out to downstream systems – no custom scripts, no maintenance burden.

And because it's a verified community node, it's vetted by n8n and installable straight from your instance.

The node supports everything our REST API does, so anything you can do outside n8n, you can now do inside it.

What you can do

Ingest

• Save – ingest or update a source entity
• Delete – remove a source entity permanently
• Bulk save – ingest or update up to 50 source entities in one call
• Bulk delete – delete up to 50 source entities in one call

Delivery

• Get content – retrieve a view by URL, ID, or handle

Query

• Query Items – query a single index
• Multi Query Items – run up to 5 queries in a single request

Route

• Get Many – retrieve a list of routes for the environment

A quick example

Imagine a nightly workflow: an n8n Schedule trigger fires → an HTTP Request node pulls the latest product from your PIM → the Enterspeed node does a Bulk save. Your Query API stays fresh, and your PIM stays offloaded – all on autopilot. 🚀

Getting started

Check out the Enterspeed n8n node and n8n's guide on installing verified community nodes. Full details are in the Enterspeed n8n documentation.

Questions? Reach out on our support Slack channel or at support@enterspeed.com.

You've been able to ask questions about your data with the Query MCP. Now you can ask about your tenant too. 🛠

The new Enterspeed Management MCP turns the Enterspeed Management API into a set of Model Context Protocol (MCP) tools that AI clients like Claude, Cursor or VS Code can discover and call on their own. Where the Query MCP reads your data, the Management MCP reads – and changes – your configuration: sources, environments, domains, clients, and the mapping schemas that tie it all together.

It's available now, so go ahead and give it a try.

Why it matters

For a long time we've had the Enterspeed CLI to read or configure a tenant from code, and while the Management MCP can do more or less the same, it's more tailored to agents. Now you can simply ask: "Give me an overview of my tenant", "What triggers the schema behind my product pages?" or "Show me any processing errors in the last 24 hours". You can even have an agent author a mapping schema, dry-run it against a real source entity, and save a new version, all in one conversation.

Read more in the Enterspeed Management MCP documentation.

Nobody types perfectly. A shopper searching your site will mistype "televisiom", drop a letter, or add one too many – and until now, that meant no results.

The Enterspeed Query API now supports fuzzy search, so small typos no longer return an empty result. Instead of demanding an exact match, it finds the closest results and ranks them by relevance – giving people what they're looking for, even when they don't spell it quite right.

It's a natural fit for search bars, autocomplete, and anywhere your users type freely. Fuzzy search works on analyzed text fields in your index and can be combined with your existing filters, sorting, pagination, and facets.

A quick example

Request – note the deliberate typo in televisiom:

POST https://query.enterspeed.com/v1/productIndex

{
  "search": {
    "field": "name",
    "value": "televisiom",
    "literal": false
  }
}

Response – the correct item still comes back, ranked by _score:

{
  "status": 0,
  "totalResults": 1,
  "results": [
  {
    "sku": "p-5427",
    "name": "Television",
    "url": "/products/television/",
    "_score": 1.389
  }
  ]
}

Try it in the Management App

Search is also available directly in the Query playground in the Management App – no code required. Build and tweak your query, hit run, and see results instantly.

And as a nice extra: once you've built your query, you can now just click the cURL button to copy the full cURL request based on your query. 🚀

Check out the full details in the Query API documentation.

Your data already lives in Enterspeed – products from your PIM, orders from your ERP, customers from your CRM. Until now, getting answers out of it meant writing queries. What if you could just… ask? 🤔

That’s exactly what the new Enterspeed Query MCP unlocks. It turns the Enterspeed Query API into a set of Model Context Protocol (MCP) tools that AI clients like Claude, Cursor or Azure AI Foundry can discover and call on their own – answering questions in plain English, grounded in your real data across multiple sources.

The Enterspeed Query MCP is now available in public preview, so go ahead and give it a try.

Why it matters

AI is only as good as the data behind it. Enterspeed already does the unglamorous part – connecting your backend systems into one clean, structured, fast data layer. The Query MCP simply opens that layer to AI clients in a governed way: scoped keys, index restrictions and read-only access, all enforced by the Query API.

So no more pinging a developer for a one-off BI query or HTTP endpoint every time you need a specific answer. Merchandisers and analysts can just ask – no system logins, no query language, only read-only access to the data already in Enterspeed.

Read more and see a concrete commerce example in the Enterspeed Query MCP blog post or checkout the documentation.

If you've been building with Enterspeed  –  it means you know better than anyone what works, what doesn't, and what's missing.

We're working on the next major version of the platform, and before we lock in direction, we want to hear from the people actually using it. 

The survey takes about 5 minutes, and your answers will directly influence what we build next.

[Take the survey →]

Oh, and everyone who completes the survey is entered into a draw for an Enterspeed hoodie. 💙 (Just leave your name at the end – we'll only use it for the draw.) 

Brutal honesty welcome 😬

We wanted to give you a heads-up that the global search feature in the Enterspeed Management App — powered by CommandAI's Spotlight — will be discontinued later this month (June 2026).

The reason is twofold: CommandAI is ending support for the Spotlight feature, and looking at our usage data, only a small number of users have been using it regularly.

We know it's a handy feature for navigating the Management App, but based on the low usage and the loss of support, we've decided not to build an alternative — at least for now.

We're sorry for any inconvenience this may cause, and as always, feel free to reach out if you have any questions.

Are you working with large number of entities, being customer specific prices or entities that changes often like e.g. inventory numbers?

Up until now Enterspeed has had its strength on static data from a CMS, a PIM system or something similar. This is because the way Enterspeed preprocesses the data making it stable and fast to fetch on the client.

But no approach is the perfect solution for all use cases, and we know that not all data is static and requires transformations. So, why put data through a transformation process that takes a lot of resources if the data doesn’t need transformation or if it’s never even being fetch by the client?

That’s why we added a new type of data flow in Enterspeed called Auto Indexing.

It’s a feature you enable on the Source group. Once it’s enabled all the data you ingest into this Source group will automatically be added directly to an Index you can query from the Query API.

The new Auto Indexing mode lives comfortably together with the existing schema transformation mode, given you the options to use whatever mode is the best for your specific data set and often it will be a combination of both for your project.

When using Auto Indexing you don’t have the option to create schemas for the source entities, so you can’t do transformation on the data. But instead, you can ingest a large volume of data and have frequently changes to the data making it perfect for dynamic data like customer specific prices or inventory numbers. And especially in combination with the bulk Ingest API you can really send some data through to Enterspeed and have a fast and flexible Query API to fetch the data and offload your ERP system or wherever your data is coming from 🚀

Head over to your Enterspeed tenant today and start using the now Auto Indexing mode and if you need more information, check out the documentation about Auto Indexing.