Configure through the Form first
Collect viewpoint, activity, result and operating conditions on the website. The Form prepares the context and the handoff for package composition.
Read the Form routeInside the MAIOS package
The project seed
The package takes the information you provide and turns it into an agentic system that understands what can be done to evolve solutions and develop the most promising ideas.
02 · Operating seed
MAIOS packageProject Kernelstate + competences
03 · Project assistant
Operating context
Objectives, sources, constraints and competences orient the work.
- The person provides the work, objective, context, sources, constraints and responsibility.
- The package organizes that information in the Project Kernel.
- The assistant uses the Project Kernel as the operating context for the work.
Inside the Project Kernel
What changes in the assistant
A general-purpose assistant knows its model and the tools available to it. With the Project Kernel, it receives an operating form for this project: it knows what must be achieved, which sources govern the work, what has already been decided, which competences to use and where to resume.
RepoKernel generates this base. MAIOS adds the startup interview, operating kernel and faculty router so the structure can become specific to the real work.
01Context and sourcesSeparates facts, decisions, hypotheses and material to verify.
02State and continuityPreserves what was decided and the point from which to continue.
03CompetencesSelects and combines methods that contribute to the result.
04Actions and checksConnects each movement to responsibility, conditions and evidence.
05Open possibilitiesDoes not reduce the project to its initial options and preserves relevant differences.
06EvolutionTurns useful corrections into reviewable, reusable capabilities.
The kernel currently prepares the project and its way of working. Tools, sensors, databases, accounts and automations enter when they are connected and authorized in the real context. The same structure can then support progressive autonomy through verified capabilities.
Who it was conceived for
Nine contexts, one system that learns the work
The package is for people and organizations that want to turn occasional AI use into a system aware of its operating context. The following cases identify work the kernel can prepare and coordinate as it receives suitable competences, tools and permissions.
01AI builder or developerAI agents, workflows and software
- Designing AI assistants and multi-agent systems.
- Orchestrating agentic workflows, APIs and tools.
- AI agent testing, evaluation, observability and monitoring.
- Technical documentation, releases and security boundaries.
02Company or leadershipProcesses, decisions and business continuity
- Business process automation and knowledge management.
- Decision support, reporting and operational control.
- Customer management, internal requests and CRM workflows.
- AI governance, company data, risk and continuity.
03Research or universityExperiments, knowledge and transfer
- Literature research and source maps.
- Experiment planning, experiment tracking and reproducibility.
- Monitoring data, instruments and laboratory cycles.
- Research project management and transfer of results.
04Professional or firmClients, documents and responsibility
- Client intake, document collection and matter organization.
- Research, analysis and draft preparation with identifiable sources.
- Deadlines, reviews, approvals and responsibility handoffs.
- Local or hybrid AI for confidential professional information.
05Consultant, startup or AI agencyDelivery, automation and clients
- AI assessment and client solution design.
- Workflow automation, CRM and marketing operations.
- Reusable delivery methods across different projects.
- Continuity of decisions, materials and responsibility.
06Manager or team leadProcesses, people and outcomes
- Process mapping, bottlenecks and dependencies.
- KPI, risk, issue and decision monitoring.
- People coordination, handoffs and responsibility.
- Project portfolio management and operational continuity.
07Operator or teamDaily work, quality and safety
- Standard operating procedures, checklists and contextual instructions.
- Request classification and routing.
- Exception management and quality or safety monitoring.
- Shift handoffs, reporting and continuous improvement.
08School, institution or organizationServices, learning and coordination
- AI training pathways and knowledge management.
- Organizing requests and services for people.
- Coordinating projects, activities and stakeholder communication.
- Governance, accessibility, privacy and responsible AI use.
09Investor or project evaluatorDue diligence, risk and milestones
- Technical due diligence for startups and AI systems.
- Comparing claims with available sources and evidence.
- Risk assessment and milestone or outcome monitoring.
- Comparable scenarios and decision-support memos.
Two entries into the same project
Choose when to provide the context
AI Setup describes both routes and what you receive. The difference is when the project is configured.
Configure inside the folder
Download a package with an initial Project Kernel. In a new folder, the assistant runs the startup interview and makes the kernel specific to the project.
Download the self-configuring packageCurrent boundary. Self-configuring package 1.2.2 starts a project in a new, empty folder. External tools, publication, monitoring and actions on real systems require their own connections, permissions and checks.