# /loop, AI products that work inside your business > /loop is Deploiable’s model for building AI products that work inside your business, from identifying what is worth building to measuring the results. Source: https://deploiable.com/loop.html /loop # We build AI products that work inside your business. /loop is Deploiable’s model for identifying what is worth building, putting it to work in your company and measuring the results. [Let’s build your AI product](https://deploiable.com/contact.html) See the path The idea ## AI is not a project that closes. It is a cycle that improves at every turn. An AI product starts creating value when your company uses it. With /loop, we build it, put it to work and keep measuring how it performs. The usual AI project Stops at delivery The usual AI project goes from pilot to demo to delivery, then it stops and the promised value remains on paper. With /loop, AI products observe, act, measure the result and start again, inside your business. A pilot, a demo, a delivery. Then it stops. /loop Always running It observes, acts, measures the result and starts again. ### Every turn is measured - against the agreed targets. ### People stay in the loop - AI works in the background; your team decides. ### Delivery is where it starts - Where the usual project stops, the loop begins: in daily use, producing results turn after turn. What you get ## Four dimensions of value. Measured, not promised. We build products around what your business needs: better information, support for decisions and tools that do the work. We measure their impact in four areas. ### Time - Less time from a question to an answer, or a need to action. ### Cost - Lower costs for the work your product supports. ### Capacity - More work your team can take on with the same resources. ### Quality - More reliable information, fewer errors and better decisions. Built for ## For SMEs and large organizations. For companies ready to put AI products to work in their business, with goals they can measure. ### SMEs AI products built around your business, at a scale that fits your company. ### Large organizations AI products integrated with your data, tools and teams, with a governed path from the first version to measured results. How it works ## Five phases. You invest only in what creates value. Each phase ends with a result your company reviews and approves before the next one begins. ### Business Understanding - We meet your teams to understand your business, its goals, the information it relies on and how work gets done. The result is a shared picture of where an AI product could make a difference. ### Priority Framework - Together with your team, we identify the AI products that could create the most value for your business and rank them in a shared priority roadmap. ### Impact Modeling - We estimate the expected economic value of each priority and set measurable targets. The result is a business case on which your company can base its investment decision. ### Solution Build - We prepare the infrastructure and build the AI product, connecting it to your data and the systems already in use so that it runs stably and securely inside your business. ### Built to Work - We release the product into production, support your team through adoption and measure the results for twelve months, with quarterly reviews that allow us to intervene as soon as results deviate. [See /loop in practice](https://deploiable.com/lab.html) How we are paid ## Our fee depends on your results. We share the risk of innovation. A significant part of our fee is a success fee, paid when the objectives agreed with you are reached. How our fee works Twelve months measured Your result rises over twelve months of measured results, with quarterly reviews. When it passes the agreed objective, the success fee applies. Agreed objective Your result Objective reached Success fee Start of the path Twelve months of measured results Above the agreed objective, the success fee applies. Questions ## What companies ask about /loop. ### How is Deploiable paid in /loop? A significant part of our fee is a success fee, paid when the objectives agreed with you are reached. Results are measured for twelve months, with quarterly reviews. ### Can we start without committing to a full project? Yes. The first three phases (Business Understanding, Priority Framework and Impact Modeling) show which AI products are worth building for your business and what value they could create. The investment is limited, and nothing gets built until you decide. ### What does /loop measure? Four dimensions of value: time, cost, capacity and quality. Every result is measured against the targets set together. ### Does AI replace our team? No. It works in the background, and your team stays in charge of every decision. Everything runs inside the tools your people already use. ### Who is /loop for? SMEs and large organizations that want AI products built for their business, integrated with their data and tools, with results they can verify. ### Why is it called /loop? A loop is how every AI agent works: it acts, observes the result and starts again. The AI products we build through /loop put that cycle to work inside your business. In Claude, /loop is the command that repeats a task at regular intervals. ## Learn where AI is worth it before anything is built. The first three phases show which AI products could create value for your business and what they are worth, with a contained investment. [Let’s build your AI product](https://deploiable.com/contact.html)