Deyel

What the product is

Deyel is a low-code platform that adds generative AI to the building of custom business applications. Its FAQ answers the question "what is Deyel" this way, in Spanish: a low-code platform that integrates AI to create tailored business applications, combining the speed of AI prototyping with the solidity an enterprise environment demands, with governance, security and compliance built in from day one. The site is written mainly in Spanish, with an English section, and the footer lists contact numbers in Argentina, Brazil, Chile, Mexico and Spain. It is aimed at IT teams and business units at companies in sectors such as insurance, banking, financial services, government, manufacturing and retail that need software shaped to their own process, and the solutions menu speaks separately to IT leaders, professional developers, business leaders and business specialists who want to build without coding.

The same entry serves a reader coming from the BPM and process-automation page, and the process side is covered below.

What you build and how

The home page frames the product as taking ideas from prototype to production with low-code and generative AI. It names 4 themes: low-code plus AI, native integration, centralized governance, and customization plus extensibility. It describes one operating environment with 2 ways of working.

Enterprise applications, built prompt-first. Applications are created through a prompt-first approach with the help of DIANA, which the site calls the AI built into development. Through natural-language prompts DIANA generates visual components inside the platform, namely forms, flows, pages, business rules and entities. The company stresses that DIANA does not produce loose code that IT later has to decipher. Its claim is that "DIANA construye dentro de Deyel" (DIANA builds inside Deyel), so what it creates is operable and auditable from the first moment, and you can keep evolving the result in a hybrid way, either by asking DIANA for changes or by adjusting components yourself in the visual modeler. During development DIANA suggests the next step, completes logic and flags inconsistencies before production. Asked whether the AI acts on its own, the FAQ answers "Depende del rol asignado" (it depends on the assigned role): in development DIANA proposes components while you keep approval control, which the company calls human-in-the-loop, and in daily operation agents can act autonomously within the governance policies and permissions the organization defines.

Agentic applications. The second mode adds AI agents to an application you already have. The company says agents evaluate context, reason, collaborate and carry out tasks from start to finish in real time, so teams move from manual data entry to what the page calls management by objectives, where users state what they need in natural language and the agents run the checks, queries and processes. A section on safe coexistence says agents work within the system's business rules, entities and permissions. Through what it calls the AI Runtime, each interaction follows the security model and existing logic, and the AI model never reaches data directly or without control. Agents are configured in the Agent Modeler, which the FAQ distinguishes from chat tools that only return text: it sets up agents that interpret context, evaluate data and carry out operational tasks on top of existing business rules and processes.

Integration. The home page calls Deyel an API-first platform designed to coexist with core systems, ERPs and external services. The FAQ says Deyel has a native integration hub that connects to legacy systems, ERPs, databases and external services through APIs and standard protocols. It describes the hub as the layer that unifies what the organization already has, with no need to replace it. For specific cases, the company says its technical team can assess compatibility in a diagnostic session.

Process automation. The navigation separates Automations, described as automating workflows to reduce unintentional errors and raise team productivity, from DeyelBPM, which is presented as business process for business users, and a process gallery. The company's English blog on banking gives concrete patterns. It says a low-code platform can orchestrate an account-opening workflow while AI handles identity verification, and that loan approvals, claims management and asset administration are workflows suited to this approach. It also describes intelligent document processing, where AI extracts and classifies information from invoices, contracts and forms and "the Low Code platform orchestrates the workflow for review and approval by the human team." Approval workflows that cross departments are listed as a candidate for automation, with AI helping prioritize tasks and assign cases.

What gets built. The FAQ names customer portals, financial approval flows and end-to-end operational management systems as typical builds. The resources menu offers templates, which the company says speed up development by 80%, and a Vendor Management product for purchasing and supplier management. The banking blog adds that chatbots can be created with AI to answer customer questions around the clock.

Working model. The FAQ says most customers start with a pilot: pick a painful, real problem, build it in Deyel in weeks, validate it in production, then expand. The collaboration model is flexible, so you can lead the work, bring the vendor in at key stages, or delegate the whole build. The navigation also offers a self-service project estimate, documentation, training, a test drive and a partner program, with a directory for finding an implementation partner.

Plans, pricing and proof

Pricing. The navigation links a pricing page, but the site shows no price figures, plan names or usage limits. The home page has two calls to action, "Comenzar Gratis" (start free) and "Agendar una Demo" (schedule a demo), and a self-service estimation tool sizes your next application.

The cost argument. The FAQ gives three ways Deyel lowers total cost of ownership. Build time falls by up to 70 to 90% compared with traditional development, maintenance stays low because there is no hand-written code to patch, and, in Spanish, "con Deyel construir una solución custom para tu proceso exacto cuesta lo mismo que adaptar un SaaS que nunca termina de calzar" (with Deyel, building a custom solution for your exact process costs the same as adapting a SaaS that never quite fits). The page calls the alternative an "impuesto a la innovación" (innovation tax) of paying for something that does not move at your pace. These are the company's own figures and claims.

Customer stories. The home page summarizes three: Banesco modernized its credit origination flow without disturbing its core systems; General de Seguros, a Mexican insurer the page says has 50 years in the market, gained visibility into its internal flows; and AON built a complete life insurance system. Roberto Lynch, claims deputy director at AON Argentina, is quoted saying Deyel provided a professional development team that let them meet a tight deadline with a fully working product. José M. Kobashigawa Kiyán, operations and IT manager at Cooperativa Pacífico, credits faster response times and more flexibility. An IT expert in insurance, named on the page as Anton López Krall, is quoted: "Con Deyel comenzamos de a poco, sin comprometer grandes recursos y escalamos de acuerdo a las necesidades de la compañía. El resultado fue un proyecto muy útil, ejecutado en tiempo y forma y con los recursos necesarios." In English, they started small without committing large resources, scaled with the company's needs, and got a useful project delivered on time.

Security, ownership and hosting. The FAQ says Deyel includes granular role and permission management, an audit log of every change, separate development and production environments, and international compliance standards, without naming a standard. The AI Runtime is its stated mechanism for keeping AI inside existing permissions. On ownership, the FAQ says the business logic, data and processes you build belong 100% to your organization, and that with the on-premise option the infrastructure and applications stay running inside your company.

Who it suits and who it does not

Deyel suits a company, often in insurance, finance or government, that has a specific process poorly served by packaged software and wants to build it with AI assistance while keeping permissions, audit and integration with legacy systems under central control. The pilot-first approach and the choice to build yourself, co-build or delegate make it workable for teams at different levels of in-house skill.

It is a poorer match for a solo founder looking for a website or storefront builder, or for a team that wants a published price list before the first conversation. The site is mainly in Spanish, so teams working only in English will rely on the English section and documentation. A regulated buyer letting agents act on business data will want to test the AI Runtime controls in the pilot before relying on them.