INDEPENDENT SOFTWARE ENGINEERINGFLORENCE, ITALY / WORKING ACROSS EUROPE

devFlorence

Dedicated applications.
Java migration.
Hands-on architecture.

A software studio led by Alessandro De Stasio. Thirty years of engineering experience, with direct responsibility for what reaches production.

Discuss your project ↗

Services

Engineering services.

devFlorence develops systems shaped by business processes and migrates established applications through controlled, incremental change. The work combines senior architecture, modern Java engineering, lean technology choices, and production-focused delivery.

01

Dedicated business applications

Own the software that runs your business. When a broad SaaS suite becomes costly and cumbersome, a focused application can fit your processes and business model, without making you reshape them around someone else's product.

AI-assisted development makes this option more practical. We build the workflows, integrations and automation you actually need, with senior engineering oversight and clear ownership of the code. Keep the services that earn their place; replace the parts that hold you back.

03

Architecture & technical leadership

System design, technical direction, trade-off analysis, and architecture that stays understandable after the first release.

04

Cloud, AI & application integration

AWS-native systems, real-time messaging and AI assistants connected to your application's workflows. Context policies reflect the use case, shareable data and budget. Identity integration includes OAuth 2.0 and Amazon Cognito.

Application migration

Java application migration.

devFlorence migrates established Java and Java EE systems with current Java, Jakarta EE, MicroProfile, Quarkus, and GraalVM. The path is incremental, dependency-conscious, and directed toward cloud or serverless deployment where it reduces operational work.

Explore the Java migration service
01

Assess

Map code, dependencies, data, integrations, domain language, and operational constraints before choosing a target architecture.

02

Protect

Recover specifications and establish characterization tests that preserve essential behavior while the system changes.

03

Migrate

Upgrade, replace, or extract one bounded part of the system using current Java and the smallest useful technology set.

04

Deploy

Release the accepted increment to cloud, serverless, or container infrastructure with observability and rollback controls.

Java migration stack

A mature ecosystem for controlled change.

Java, Java EE, Jakarta EE, MicroProfile, Quarkus, and GraalVM provide stable contracts, explicit types, mature tests, compact runtimes, and practical deployment options from containers to AWS Lambda.

AI-assisted analysis can reduce mechanical work during discovery and migration, while engineering review remains authoritative. The approach is informed by pragmatic, standards-first Java practices and recent Airhacks work on spec-driven Java development with LLMs .

Principles

The way we build.

devFlorence keeps the engineering surface intentionally focused: clear process models, explicit boundaries, modern Java foundations, and AI assistance under senior technical control.

KISS, YAGNI, DRYPrefer simple designs that solve today’s process requirements. Avoid speculative features, duplicated logic, and abstractions without evidence.
Process model firstEncode the business process clearly: rules, states, decisions, data ownership, and the language your team already uses.
ECB architectureOne of the most effective architectural patterns for Java applications, Entity–Control–Boundary separates domain state, use-case coordination, and external interfaces. It keeps business logic explicit and testable while giving LLMs a clear, understandable structure for AI-assisted development.
Production‑ready architectureChoose boring where possible and modern where useful: Java, serverless, reactive patterns, observability, and maintainable boundaries.
AI-assisted developmentAI assists analysis, implementation and testing. Architecture, security and production decisions remain the engineer's responsibility.

Featured product

Diaktor

Our own product. The same engineering.

Diaktor is a messaging platform deployed in the customer's AWS account, with Matrix-compatible E2EE, real-time push delivery and standard Matrix federation. Cognito provides integrated account access. Diaktor AI integrates an assistant into rooms and direct chats.

Diaktor AI supports Amazon Bedrock and OpenRouter, with provider and model configured per project. Context policies are tailored per project. The demo uses the current request, while memory per room and conversation summaries require development. Content selected for AI is processed by the configured provider outside the room's E2EE boundary.

A demo is available. The beta programme is reserved for the first five participating organizations planning a rollout to at least 500 Diaktor users, starting with a smaller beta group. Participants receive free adoption analysis and consulting, scheduled guidance, and 15% off agreed custom Diaktor AI development and other Diaktor extensions during the first three months. Explore early access ↗

Demo available · Beta programme

Secure end-to-end encrypted messaging, Matrix-compatible and built to scale.

www.diaktor.net
Secure by designMatrix-compatible E2EE protects message content and attachments between participating clients.
Elastic AWS operationRuns in the customer's AWS account, scales with demand, and delivers updates through real-time push events.
Identity and Diaktor AICognito-backed account access and Diaktor AI are integrated with the federated messaging core. The current AI worker uses EC2.
01 Architecture under pressure

Federation, message flow, event propagation, AWS services, and reliability trade-offs.

02 Application integration

AI provider integration, context policies, Cognito authentication and access boundaries within one product.

03 The devFlorence method

Senior judgment, focused scope, lean delivery, and AI-assisted execution applied to a real product.

Alessandro De Stasio with devFlorence and Diaktor studio branding

Alessandro De Stasio

Software architect, developer, and founder of devFlorence.

About

One technical owner, from first discussion to delivery.

devFlorence is a senior-led software studio founded by Alessandro De Stasio, bringing 30+ years of software engineering experience and 28 years on the Java technology stack.

Trusted AI agents extend research, implementation, testing, documentation, and delivery speed — but architecture, trade-offs, code quality, and production decisions remain senior engineering responsibilities.

Alessandro's experience includes working in an international team led by Adam Bien and continuing professional development through Airhacks workshops since 2018, including the 2026 course on spec-driven Java development with LLMs.

Compared with a traditional agency, the model is lighter and more direct: fewer layers, faster decisions, and continuous technical ownership grounded in the process, the architecture, and the long-term cost of the software.

30+years software engineering
28years Java stack
1:1direct senior collaboration

Contact

Let's talk about your software.

Tell me about the application, legacy system, AI integration or process improvement you have in mind. devFlorence can help define the use case, identity and data boundaries, then build the software your business needs.

Your information is used only to review and reply to this request. You can request access, correction, or deletion through the privacy notice.