Senti
Industries / Forex VPS, Hosting & Algo-Trading Infrastructure Providers
Sell more than the machine.
Senti's automation layer, installed on your fleet and delivered from your admin system under your brand. Your infrastructure, customers, and billing stay yours.
These figures show platform activity and availability. They are not investment performance or return claims.
Not a competitor. The layer above your product.
It would be easy to read the rest of this site as an argument against VPS hosting. It is not. Senti has never removed the VPS layer: MT5 requires a Windows terminal on server infrastructure, and that is as true inside Senti as it is on your fleet.
What Senti does is abstract that layer. The end user stops provisioning environments, keeping terminals connected, deploying EAs by hand, and checking on them over RDP. That abstraction is the product, and it is separable from whose infrastructure runs underneath it. In Senti’s own channel, Senti operates the infrastructure and the user works under Senti’s brand. In this channel, you operate the infrastructure, and your user works under your brand, in your admin system.
Same operating principle. Different distribution channel to the end user.
Your product stops at the machine
You deliver the infrastructure layer well. The layer above it stays with the customer, and that is where the recurring load comes from.
- The customer does the operating work. Provisioning the terminal, installing MT5, attaching the EA, enabling algo trading, and reconnecting after a restart all remain theirs.
- Remote desktop is the interface, so every routine check requires a client, credentials, and a workable connection, and is awkward from a phone.
- Terminal death is silent. A disconnected MT5, a crashed EA, or a failed login often surfaces only when the customer notices missing trades.
- Support tickets are mostly status questions, not infrastructure faults, so the desk absorbs work that a status surface would have answered.
- Instances look identical from the outside. Uptime on the box does not tell you whether the thing the customer actually pays for is running.
- The offer is hard to differentiate. Specs, location, and price are comparable across the market, so competition compresses toward price.
None of this is a fault in the hosting product. It is the gap between “the server is up” and “the customer’s automation is working”, and closing that gap is exactly what the abstraction layer does.
What Senti Inside does
Your team installs the Senti Inside agent onto the MT5 terminals running on your instances. The agent automates the operating work the customer is doing by hand today and reports the state of each terminal to a surface your customers reach from a browser. The intended scope, subject to technical confirmation:
- Terminal state — whether MT5 is running, connected to the broker, and logged in to the expected account.
- EA state — which Expert Advisors are attached, whether algorithmic trading is enabled, and whether the EA is active.
- Account state — balance, equity, open positions, and open orders for the connected account.
- Activity record — timestamped order activity from that terminal, so a support conversation can start from evidence instead of a screenshot.
- Control actions — the supported operations on a managed terminal. The exact set, its permission model, and its failure behavior are confirmed during technical scoping and must not be treated as an arbitrary remote command channel.
- Health signals — notification when a managed terminal stops reporting.
Your instances keep running the terminals, exactly as they do now. Senti supplies the layer that operates them.
The only two things that change
The operating principle is the same one Senti runs in its own product. A partner conversation goes wrong when this is presented as a different architecture, so be precise about what actually differs.
Two rows move: whose infrastructure it runs on, and whose product it arrives in. Everything else is identical. This is also why Senti Inside is not a migration path onto Senti’s infrastructure. Your fleet keeps running your customers’ terminals, and if a partner conversation needs that written into the agreement, it should be.
The channel runs both ways
The obvious direction is that you distribute Senti. The less obvious one is that Senti distributes you, and a partner agreement can carry both.
- You distribute Senti. Your customers get automated MT5 operation inside your admin system under your brand. They never need to know Senti is underneath.
- Senti distributes you. Senti’s own users need infrastructure beneath their automation, and the entire point of the product is that they never choose, configure, or think about it. That is demand for hosting capacity where the end user has no vendor preference to satisfy.
Each side is invisible inside the other’s product, which is what makes both directions work at once.
What each side owns
Why a hosting provider would want this
- Move up from instance rental. The offer becomes a managed automated service rather than a machine with a specification sheet.
- Sell a difference that is not specs or price. An automation layer is a product attribute competitors renting raw instances do not have.
- Answer status questions without a session. The most common support ticket becomes something the customer resolves themselves.
- Give the sales team a demonstrable feature. A surface can be shown; a latency figure has to be believed.
- Detect silent failure earlier. A terminal that stops reporting is visible before the customer discovers it through missing trades.
- Support a higher service tier. The layer gives a concrete reason for a managed plan to sit above a bare-instance plan.
- Reach customers who are not at a desk. A browser surface works from a phone in a way a remote desktop client does not.
These are operational and commercial outcomes for your hosting business. Senti makes no claim about your customers’ trading results, and the addon must never be positioned to end users as improving trading performance.
Fit and qualification
That last point is a genuine gating question, not a formality. Deploying an agent onto instances your customers control has terms-of-service, consent, and data-handling implications that differ by jurisdiction and by how your contracts are written. Bring your terms to the first conversation.
No commercial model is confirmed
The rest of Senti’s B2B services bill per provisioned Trade Account, because a Trade Account is the unit of Senti’s compute cost. That basis does not carry over here: your infrastructure runs the terminals, so Senti’s cost shape is different.
The reciprocal structure above complicates it further, in a way worth stating plainly rather than papering over. If an agreement carries both directions, the question is not a unit price at all. It is how a layer-for-capacity exchange settles between two parties who are each other’s supplier and channel.
This page publishes no rate, no per-terminal price, no revenue share, and no setup fee. The billing shape depends on whether an agreement carries one direction or both, and that is scoped in the partnership conversation rather than quoted from a rate card.
How an engagement would start
Walk through your hosting architecture, instance images, control panel, customer terms, and the support load you want to reduce.
Agree how the agent reaches the terminals: baked into your instance image, deployed through your existing management tooling, or installed per customer on activation.
Settle the data scope, permission model, supported control actions, authentication, branding, and the support handoff between your desk and Senti’s escalation path.
Run the addon on a small, named set of instances with agreed acceptance criteria before it reaches the general customer base.
Decide which plans include it, how it is presented to customers, and what your support team is trained to answer.
Extend to the wider instance population once the operational model holds.
Timelines depend on your instance management, panel architecture, and terms review. They are confirmed after the technical review rather than promised in advance.
Frequently asked questions
Does this move our customers onto Senti’s infrastructure? +
Is Senti not a competitor to VPS providers? +
What can a customer actually do from the surface? +
Does Senti see our customers’ strategies? +
Bring your fleet. Leave with a scope.
Bring your architecture, control panel, customer terms, and the tickets you would most like to stop answering. We will scope the layer and name the open questions.