CFOtech Ireland - Technology news for CFOs & financial decision-makers
Ireland
Graftcode moves integration platform to stable release

Graftcode moves integration platform to stable release

Thu, 19th Nov 2026 (Today)
Joseph Gabriel Lagonsin
JOSEPH GABRIEL LAGONSIN News Editor

Graftcode has moved its software integration platform to a stable release, marking its shift from alpha and beta development to production use.

The Warsaw-based company says the release adds authentication, authorisation, OpenTelemetry-based observability tools, asynchronous calls, and streaming and messaging integrations. It also adds support for the Model Context Protocol, or MCP, and development rules for AI coding tools including Cursor, VS Code and Claude Code.

The platform is designed to reduce the amount of integration code needed when applications, services and front ends communicate across systems and programming languages. Developers expose existing application methods through Graftcode Gateway, while consuming applications install automatically generated, strongly typed dependencies through standard package managers.

Graftcode raised €2.1 million in April. The stable release follows work during its alpha and beta phases to validate the architecture, expand compatibility and gather developer feedback.

AI coding

The release comes as software groups weigh whether AI coding tools improve productivity when they also generate more code to review and maintain. Graftcode argues that software integration is one source of that extra code, as agents produce controllers, APIs, client SDKs and related layers that need updating when underlying systems change.

Research cited by the company points to the scale of the issue. Forrester Research has estimated that about 70% of application development work goes into glue code and wiring, while a 2026 analysis of 623 million code changes found that heavy AI users produced 25% more code, code duplication rose 81% and refactoring declined.

Rather than generating integration layers more quickly, Graftcode says its approach is to remove some of the need for them. Backend functionality can be reused across services and front-end applications, and exposed to AI agents as MCP tools, without building a separate MCP server.

Support for AI coding agents is a central part of the stable release. The new development rules are intended to help coding agents understand how to build systems using Graftcode within AI-assisted development workflows.

Benchmarks

In its own tests, Graftcode says the platform reduced integration code by about 30% to 60% in applicable scenarios. It also says AI-assisted development tests showed about 30% to 60% lower token usage.

Selected runtime tests showed the platform performing up to about 70% faster and using around one-eighth of the CPU compared with REST-based implementations, according to the company. Graftcode has made benchmark code and demos publicly available so developers can inspect and reproduce the tests.

The stable release also broadens support for programming structures and introduces an open-source plugin architecture for additional transports and messaging systems, including RabbitMQ, Kafka, Azure Service Bus, Azure Event Hubs and AWS SQS.

Other additions include a streamlined command-line interface, support for CORS and HTTP/2, and documentation covering supported production scenarios, type combinations and known limitations.

Enterprise use

Graftcode also highlighted a proof of concept carried out by Impleflow, a Polish IBM Business Partner working on IT service management software including IBM Maximo. In that project, Impleflow used Graftcode to integrate Maximo with other systems for an enterprise client.

The implementation used Graftcode Gateway to expose selected methods from Maximo deployment-specific classes. According to the company, that made existing business functions available to mobile applications through strongly typed Grafts and to AI agents through MCP, while preserving existing authentication, authorisation and observability mechanisms.

Impleflow says it used the system to build mobile interfaces and an AI assistant that can retrieve, create and update Maximo records.

"Developers have spent years building controllers, DTOs, clients and other glue code just to make applications communicate. With AI-assisted development, that cost becomes even more visible because AI agents also have to generate, understand and maintain all of that integration code," said Przemysław Ładyński, Chief Executive Officer and Co-Founder of Graftcode.

"With Graftcode reaching stable release, we want to offer a production-ready alternative, letting developers and AI agents focus on business logic while decoupling projects from integration-specific code," Ładyński said.

"In our daily work we integrate IBM Maximo with many different systems, and we know how much time goes into building and maintaining the integration layer. As enthusiasts of new technology, we chose Graftcode and were impressed by how quickly a working integration could be built with it. The technology proved especially valuable at the PoC stage, where fast delivery and demonstrating value early matter most. Graftcode shortens preparation time and, as a result, reduces cost. We look forward to further PoCs and wider adoption of this technology, as we believe it will significantly accelerate integration between enterprise-class systems," said Bartosz Gancarz, Chief Executive Officer of Impleflow.