AI Solutions Architect · Forward Deployed Engineer · Python Developer · Odoo Solution Architect · Open Source Contributor

I design and build enterprise software and AI-powered solutions, connecting business processes, open-source technologies, and intelligent automation.

My work spans Python engineering, Odoo architecture, Brazilian localization, LLMs, agentic AI systems, and enterprise integrations.

I'm a co-founder of Escodoo and Deepze.ai, and an active contributor to the Odoo Community Association (OCA).

I'm a software engineer, entrepreneur, and open-source contributor focused on bridging the gap between business challenges and technology solutions.

My expertise combines enterprise software development, solution architecture, business process understanding, and practical AI engineering.

I believe the real value of technology comes from solving meaningful problems, building maintainable systems, and empowering people through open collaboration.

• 🏗️ Solution Architecture: Designing scalable, maintainable enterprise systems.
• 🐍 Software Engineering: Building robust solutions with Python, Odoo, and open-source technologies.
• 🤖 Applied AI: Developing AI agents, LLM-powered applications, RAG systems, and intelligent workflows.
• 🔗 Enterprise Integration: Connecting AI capabilities with existing business systems and processes.
• 🌍 Open Source: Contributing to collaborative software development and community-driven innovation.
• 🚀 Entrepreneurship: Transforming technical ideas into practical products and business solutions.

from dataclasses import dataclass, field


@dataclass
class MarcelSavegnago:
    """Engineer, architect, entrepreneur, and open-source contributor."""

    roles: tuple[str, ...] = (
        "AI Solutions Architect",
        "Forward Deployed Engineer",
        "Python Developer",
        "Odoo Solution Architect",
        "Open Source Contributor",
        "Entrepreneur",
    )

    companies: tuple[str, ...] = (
        "Escodoo",
        "Deepze.ai",
    )

    community: str = "Odoo Community Association (OCA)"

    psc_repositories: tuple[str, ...] = (
        "l10n-brazil",
        "fleet",
        "helpdesk-mgmt",
    )

    expertise: dict[str, tuple[str, ...]] = field(
        default_factory=lambda: {
            "software_engineering": (
                "Python",
                "Odoo",
                "Enterprise Applications",
                "API Development",
                "System Integration",
                "Software Architecture",
            ),
            "ai_engineering": (
                "Large Language Models (LLMs)",
                "Agentic AI",
                "AI Agents & Multi-Agent Systems",
                "Retrieval-Augmented Generation (RAG)",
                "Tool Calling & Orchestration",
                "AI-Assisted Software Development",
                "Enterprise AI Integration",
            ),
            "odoo": (
                "ERP Solution Architecture",
                "Odoo Development & Customization",
                "OCA Contributions & Code Review",
                "Brazilian Localization",
                "Electronic Fiscal Documents",
                "Accounting & Tax Compliance",
            ),
            "infrastructure": (
                "Linux",
                "Docker",
                "Doodba",
                "GitHub Actions",
                "CI/CD",
                "Cloud Infrastructure",
            ),
        }
    )

    current_focus: tuple[str, ...] = (
        "AI-native enterprise applications",
        "AI agents for business automation",
        "Open-source LLM infrastructure",
        "AI-assisted software engineering",
        "Odoo scalability and architecture",
        "Brazilian tax reform",
    )

    def engineering_philosophy(self) -> str:
        return (
            "Start simple, prototype quickly, "
            "validate continuously, improve iteratively, "
            "and deliver maintainable solutions."
        )

    def mission(self) -> str:
        return (
            "Bridge business and technology through "
            "open-source software, intelligent systems, "
            "and practical innovation."
        )

My focus is on Applied AI: moving beyond demonstrations and prototypes to build AI-powered solutions that integrate with real business systems.

I'm particularly interested in the engineering challenges of making AI agents useful, reliable, and maintainable.

Areas of focus:

• AI Agents & Agentic Workflows: Autonomous and semi-autonomous agents, tool calling, orchestration, and multi-step workflows.
• LLM Integration: Connecting language models with enterprise applications, APIs, databases, and business processes.
• Retrieval-Augmented Generation (RAG): Building context-aware AI applications using knowledge retrieval and embeddings.
• AI-Assisted Development: Exploring coding agents, development harnesses, iterative engineering, and AI-powered development workflows.
• Local & Open LLMs: Model serving, inference infrastructure, Ollama, Open WebUI, and open-source AI ecosystems.
• Enterprise AI Architecture: Designing solutions that combine business rules, traditional software engineering, and generative AI.

LLMs · AI Agents · RAG · Ollama · Open WebUI · Tool Calling · Agent Orchestration · Embeddings · Python

Odoo has been a central part of my professional journey.

I work with ERP architecture, custom development, integration, Brazilian localization, and complex business processes.

My approach starts with understanding the business domain and identifying how existing Odoo and OCA capabilities can solve the problem before introducing custom development.

Areas of expertise:

• Odoo solution architecture, development, and customization.
• Enterprise application integration and business process automation.
• Brazilian accounting, fiscal requirements, and localization.
• Electronic fiscal documents: NF-e, NFS-e, CT-e, and MDF-e.
• Brazilian tax reform, including IBS, CBS, and DeRE.
• Odoo deployment, infrastructure, scalability, and performance.
• Reusable modules, clean architecture, automated tests, and maintainability.

I'm an active contributor to the Odoo Community Association (OCA), collaborating with developers around the world to improve and maintain open-source Odoo modules.

I participate in the Project Steering Committees (PSCs) of:

• 🇧🇷 OCA/l10n-brazil — Brazilian localization.
• 🚗 OCA/fleet — Fleet management.
• 🛠️ OCA/helpdesk — Helpdesk and service management.

My contributions include software development, pull request reviews, technical discussions, issue investigation, and helping maintain the quality and sustainability of community modules.

I strongly believe in upstream-first development, knowledge sharing, and building solutions that benefit the wider community.

Escodoo focuses on delivering Odoo-based business solutions through consulting, implementation, customization, training, and managed infrastructure.

Our work combines technical expertise, business process understanding, and a strong commitment to open-source software.

Deepze.ai represents my involvement in AI and technological innovation, exploring opportunities to apply intelligent systems to real-world challenges.

I'm particularly interested in how AI can augment software engineering, transform enterprise applications, and create new ways for people and businesses to interact with technology.

Technology is constantly evolving, and I enjoy exploring ideas that challenge traditional software development approaches.

Some topics currently capturing my attention:

• Agentic Software Engineering: How AI agents can participate in the software development lifecycle.
• AI-Native ERP: Bringing intelligent automation and natural language interfaces into enterprise systems.
• Local AI Infrastructure: Running and orchestrating open-source models with practical deployment architectures.
• AI Development Harnesses: Tools, environments, evaluation, and workflows for more capable coding agents.
• Multi-Agent Architectures: Collaboration between specialized agents and external tools.
• Iterative AI-Assisted Development: Rapid prototyping, continuous validation, and incremental delivery.
• Open-Source Sustainability: Improving collaboration, maintainability, and contributor experience.
• Enterprise Software Architecture: Designing systems that evolve with business requirements.

Prototype → Validate → Improve → Deliver → Repeat

I believe great software is built through continuous learning, experimentation, and iteration.

My approach combines:

• Understanding the real business problem before choosing technology.
• Building small, functional prototypes to validate ideas.
• Using AI to accelerate engineering without sacrificing quality.
• Favoring simple, maintainable solutions over unnecessary complexity.
• Applying automated testing and code reviews to improve reliability.
• Sharing knowledge and contributing improvements upstream whenever possible.

AI accelerates development. Engineering judgment makes it sustainable.

I'm always interested in exchanging ideas and collaborating on:

Open Source · Odoo · Python · Applied AI · Agentic Systems · Enterprise Architecture

Building practical solutions at the intersection of software engineering, artificial intelligence, and open source.