Platform
MCP Server Build
A production MCP server for one of your systems, with up to 8 tools, secure sign-in, action logging and tests in AI…
Platform
AI Agent Knowledge Layer
A shared retrieval layer over your documents that every agent queries the same way, kept current and access-controlled.
When each AI agent keeps its own copy of your documents, answers drift apart and stale information lingers. AI Agent Knowledge Layer gives all your agents one shared retrieval layer over your knowledge base, so they look things up the same way. Access rules decide which agent can see which documents. An update process keeps the index current as files change.
- Every agent retrieves from the same current source
- Per-agent access rules on which documents each can see
- Index updates as your source documents change
$4,500
20–30 working days
Optional maintenance $249/mo
What it is
AI Agent Knowledge Layer is a build that gives your AI agents structured, shared access to your organization's documents and knowledge base, retrieved consistently rather than each agent maintaining its own copy or search logic. Access rules are set per agent, and an update process keeps the index current.
Why businesses buy it
Teams that build several agents often end up with several separate document indexes, each slightly out of date and each with its own search logic. Answers disagree, fixes have to be made in many places, and sensitive files leak into agents that should not see them. After this build, one shared layer serves every agent, with access controlled and updates flowing through. The Scope Card sets the agents, documents and controls in writing before you order, so the scope is clear and nothing is left open-ended.
Problems this solves
Each AI agent keeps its own copy of company documents
The Knowledge Layer gives agents shared, structured access to your documents and knowledge base. Agents retrieve from one source rather than each holding its own copy.
Agents give different answers because their search logic differs
Retrieval works the same way for every connected agent, across up to three AI applications. The same question draws on the same material, whichever agent happens to answer it.
An agent can reach documents it should never use
Access rules are set per agent, as up to five controls. An update process keeps the index current as documents are added, edited or retired.
A good fit if
You have more than one agent that needs to reference the same internal documents, and each currently handles retrieval differently.
Your knowledge base changes regularly and you want agents pulling from the current version, not a stale copy baked in.
You need access control on what each agent can retrieve, not one knowledge pool every agent can query freely.
You have a named technical owner who can supply cloud access and review the knowledge layer before launch.
Not the right fit if
You need agents to retain their own session history and conversation context, not shared document retrieval. See AI Agent Memory System.
Your priority is deciding which model handles a request, not knowledge retrieval. See AI Model Router.
You need the full integration layer for routing requests across providers, not a retrieval layer. See AI Gateway.
Your priority is versioning the prompts agents use, not what documents they can reference. See AI Prompt Management.
How it works
- We inventory the documents and knowledge sources your agents currently reference, including where retrieval logic is duplicated today.
- Your documents are indexed into a shared retrieval layer, up to 20 reference documents in the standard build's policy set.
- Access rules are configured, so each agent retrieves only the documents it's permitted to reference, not the full knowledge base by default.
- Agents are connected to the shared layer through one consistent retrieval interface, replacing whatever ad hoc logic existed before.
- An update process is set up so changes to source documents propagate to what agents retrieve, rather than going stale.
- A control test log documents retrieval accuracy against real queries, reviewed with your technical owner before launch.
What you provide within 72 hours
Your portal lists these the moment you order. Delivery starts the working day after everything is in and your access is checked.
- Written authorization confirming which agents and documents are in scope for the shared knowledge layer.
- A list of the AI applications and agents that should retrieve from the knowledge layer.
- Your documents and knowledge sources, up to 20, for the initial index.
- Cloud or hosting access needed to deploy the knowledge layer alongside your existing infrastructure.
- A named technical owner who can review access rules and sign off before launch.
- Admin access and API keys for every connected system, shared by adding our team as a user or creating a key you control, never by email.
- Your business information (company basics, hours and brand voice), shared during intake, and one decision-maker who can approve the final configuration.
The 72-hour rule. Send everything on your checklist, including access, within 72 hours of ordering so your build starts on time.
Timeline
Start
Intake verified, agents and documents in scope confirmed.
Weeks 1–2
Documents indexed and the shared retrieval layer built.
Week 3
Agents connected to the layer and access rules configured.
Weeks 3–4
Control test log produced and shared with your technical owner.
Week 4
Your review and two rounds of changes applied to the configuration.
Go-live
Live, recorded handoff session, 30 days of support begins.
| Option | Delivery time | Price |
|---|---|---|
| Standard | 20–30 working days | $4,500 |
| Express | 15–25 working days | $5,300 |
| Rush | 10–20 working days | $6,100 |
What you receive at handover
- AI Agent Knowledge Layer configured and deployed, serving up to three agents or AI applications
- A shared index of your documents and knowledge base, with an update process
- Per-agent access rules (up to five controls configured)
- A test report with a control test log for the agreed scenarios
- An admin guide and a recorded 60-minute handoff session
- 30 days of post-launch support
Expected benefits and how to measure them
- Can make agent answers more consistent across tools.
- Helps keep agents working from current documents.
- Makes it simpler to add a new agent without rebuilding search.
- Answer consistency
- Ask each agent the same set of questions before and after launch and compare the answers.
- Answers based on outdated documents
- Spot-check a monthly sample of answers against current document versions.
- Time to connect a new agent
- Record how long the next agent takes to connect, compared with past builds.
Before you start: List the agents in use and where each gets its documents. Write a set of test questions and record each agent's current answers.
These are things to track, not promised results. Outcomes depend on your data, your team and your market.
Running costs
The build price is one-time. Running the Knowledge Layer afterward costs roughly $20 to $300 a month in your own accounts, covering hosting, logging and monitoring tools, and it scales with how many documents are indexed and how often agents query them.
AI model usage during the build and testing is billed to your own accounts too, so every provider relationship stays yours, not ours.
Ask about running costsCustom requirements
Need more than 20 documents indexed, more than three systems in scope, or a second admin console? Add a note at checkout and you get a written Change Quote, only starting once you accept it. Larger or bespoke projects go through the Custom AI request form instead.
Request a custom versionKeep it running
For this build we recommend monthly maintenance, covering monitoring, fixes, and a monthly allowance of change hours, enough to re-index updated documents or adjust an access rule. It starts once the 30 days of support end and can be cancelled any month.
$249/mo (the minimum fee, includes 2 change hours)
Optional. Starts after the post-launch support period. Cancel any month.
How maintenance worksHow this compares
| Do it yourself with a SaaS tool | Hire a custom agency | Upmark Labs | |
|---|---|---|---|
| Price | A retrieval platform charges a monthly fee per document volume or per seat | A systems integrator quotes per project after a discovery phase | Ours is one fixed price, shown on this page before you buy |
| Scope | A retrieval platform is infrastructure your engineers configure and index themselves | An integrator agrees scope during the project | Ours is written on the Scope Card before purchase |
| Access control | A retrieval platform ships default permissions you adjust yourself | An integrator build depends on who is staffed on it | Ours is configured against the access rules you define per agent |
| Delivery | A retrieval platform is live once your team finishes indexing and integrating it | An integrator usually takes several weeks to staff and start | Ours has a fixed delivery time shown before you buy, with faster options |
Questions about this product
How much does AI Agent Knowledge Layer cost?
The fixed price for each delivery speed is shown in the price and delivery panel on this page before you order. It covers everything on the Scope Card: indexing up to 20 documents, access rules for up to three agents or applications, the update process and the control test log you review against real queries before launch.
How many documents can AI Agent Knowledge Layer index?
The standard build indexes up to 20 reference documents into the shared retrieval layer, matching your policy set for the build. More documents are added through the update process configured during the build, or through the monthly change hours that come with maintenance once the initial index is proven against real agent queries.
What does AI Agent Knowledge Layer cost to run each month?
Roughly $20 to $300 a month in your own accounts, covering hosting, logging and monitoring tools such as Langfuse or Grafana. You pay providers directly, so there's no markup from us. A small indexed set with light query volume sits toward the bottom of that range, a large index toward the top.
How is AI Agent Knowledge Layer different from AI Agent Memory System?
Knowledge Layer gives agents structured access to your organization's documents, reference material shared across agents. AI Agent Memory System stores an agent's own session history and conversation context instead. The two work well together: memory for what happened with a user, and the Knowledge Layer for what your organization knows.
Can different agents have different access to the knowledge base?
Yes. Access rules are configured per agent during the build, so an agent retrieves only the documents it's permitted to reference rather than the entire knowledge base by default. This is tested against real queries before launch and documented in the control test log your technical owner reviews.
Does the knowledge layer stay current as our documents change?
An update process is set up during the build so changes to source documents propagate to what agents retrieve, rather than agents working from a stale copy indexed once at launch. How often the index refreshes is agreed during intake, based on how frequently your documents actually change.
What do we need to provide before AI Agent Knowledge Layer can launch?
Written authorization for the agents and documents in scope, a list of the AI applications that should retrieve from the layer, and up to 20 documents for the initial index. We also need cloud access, a named technical owner, and admin access shared by adding our team as a user or creating a key you control, never sent by email.
Can I cancel maintenance for AI Agent Knowledge Layer?
Yes, any month. Maintenance starts after the 30 days of post-launch support end and runs month to month with no cancellation fee or lock-in. It includes a monthly allowance of change hours that scales with the build, and unused hours in a given month don't carry over to the next one.
Give every agent the same current knowledge, not its own stale copy.
Order AI Agent Knowledge Layer and your agents share one access-controlled retrieval layer on a fixed delivery date. The scope above is what you get, in writing, before you pay.
Works well with
Platform
AI Agent Orchestration System
Coordinates multiple agents, tools and handoffs in one runtime.
