A chronological view of the ideas, products and research directions developed by Cosen Lab.
In-depth architectural papers, engineering studies, and systemic analyses produced by Cosen Lab.
How mission-critical IBM AS/400 systems are connected to the IRIS Core AI cognitive layer, adding reasoning and traceability without replacing what already works.
AI that protects margin in every negotiation across B2B and multi-branch networks: dynamic recommendations, below-cost blocks and governed discounts.
Balance stock across the network, cut ineffective shuttle transfers by up to 30% and prevent stock-outs, with every move approved by a manager.
Track B2B market price changes, anticipate supplier dynamics and keep your price list competitive without eroding margin.
IRIS, Cognisphere, Clabbit, GeKo, Logistics Twin, Sentinel, and related projects are brought together in a portfolio that documents an ongoing, evolving program of technology research.
The protocol is further defined as a potential infrastructure for turning operational evidence into verifiable digital trust.
The project explores a new approach to cybersecurity based on continuous observation, understanding, and response.
A new line of research begins, focused on the intelligent defense of systems, data, infrastructure, and AI environments.
Governance, context, decision, evidence, and action begin to be treated as parts of a single infrastructure.
An initial experimental environment is outlined for exploring specialized cognitive systems.
AI research shifts from building a single model toward cognitive architectures composed of multiple capabilities.
The projects begin to converge on a vision in which governance, context, memory, decision-making, assistance, and security can operate as parts of a shared infrastructure.
The projects begin to converge on a common architectural vision.
Cognisphere is explored as a cross-cutting cognitive infrastructure connecting memory, context, documents, sensors, and history within E-Terna environments.
The project is further developed as an infrastructure for memory, context, and cognitive continuity.
A pilot scenario is defined for a multi-site distribution network.
The pilot moves from supporting individual operators to understanding the logistics network as a whole.
E-Terna evolves into a connected care environment in which sensors, context, memory, and interaction maintain continuity among the individual, the home, family members, and care providers.
A reflection on the future of legacy systems: not necessarily to be replaced, but to be made more intelligent.
A prototype demonstrates how intelligence can be added to a legacy system without altering how it operates.
A new interface takes shape, bringing together network status, forecasts, and preventive actions.
An architecture is outlined that can connect identities, delegations, governance, documents, compliance, evidence, workflows, agents, and AI systems within a coordinated environment.
A deterministic model is tested for assessing capacity, load, movements, and potential logistics interventions.
The concept takes shape of a digital organization able to coordinate administrative, regulatory, and operational services through IRIS, Discovery, Trust, Cognisphere, Clabbit, and Digital Twin.
The Logistics Twin evolves from a snapshot of the network to the simulation of future scenarios.
The research extends to the home environment through Digital Twin, IoT, building systems, and household documents and data, with the goal of building a persistent digital representation of the living environment.
The Digital Twin extends into logistics and distribution.
IRIS is applied to scenarios in which legacy systems and new AI capabilities must coexist rather than replace one another.
The research explores a possible evolution of the DPP toward a broader digital trust infrastructure.
The research explores how events, documents, parties, and processes can be linked through a continuous structure of evidence.
The continuous evidence model is extended to relationships between organizations and across supply chains.
The Digital Twin evolves from representing the organization to observing it over time.
Governance, identity, evidence, and trust begin to converge on a common infrastructure vision.
Discovery is extended to the structured assessment of systems and processes that involve artificial intelligence.
Simulation becomes a tool for exploring scenarios before intervening on real systems.
A dedicated environment is established for simulating scenarios and exploring the consequences of decisions.
The project is defined as a conversational operating environment rather than a simple chat application.
Clabbit evolves toward a contextual interaction model in which conversation, memory, and permissions become part of the same environment.
A reflection on the gap between the availability of information and a genuine understanding of context.
A cognitive infrastructure takes shape, designed to organize context before it is used by AI.
A new line of research begins, focused on managing context, memory, and informational relationships.
The research extends from model governance to the governance of AI systems capable of acting autonomously.
The concept takes shape of a protocol dedicated to the continuous collection and verification of evidence.
The research introduces the idea of continuous verification, rather than a periodic snapshot of an organization.
Collected evidence is organized into a reusable digital dossier for assessment and compliance.
The project evolves into a practical platform for discovery, assessment, and structured evidence collection.
A first parametric engine is developed for the structured analysis of organizations.
Organizational analysis evolves toward a deeper reading of structure, risks, processes, and dependencies.
A digital model takes shape that can represent an organization's structure, processes, data, and evidence.
The concept of AI infrastructure is extended to the digital representation of the organization.
The research moves toward a model in which compliance can be observed during day-to-day operations.
The idea emerges of representing trustworthiness through available evidence and its continuity over time.
A reflection on the relationship between automation, accountability, and the ability to reconstruct what a system has done.
Some of the first structural elements of the IRIS architecture are defined.
IRIS takes on a broader scope: not only AI, but the governance of systems, decisions, processes, and evidence.
Operational data begins to be treated not merely as information, but as potentially verifiable evidence.
The non-invasive approach is defined: adding intelligence and governance without replacing existing infrastructure.
The concept takes shape of a verifiable trail of AI-assisted decisions.
The research shifts from AI as a question-answering tool to AI as a component of decision-making processes.
Research begins on an infrastructure capable of connecting artificial intelligence, organizational processes, and decisions.
A progression from point capabilities to an interconnected digital trust and cognitive fabric.
“From AI governance and evidence to cognitive systems, autonomous intelligence and digital defense.”
From early conceptual notes in January 2025 to operational platforms, multi-domain digital twins, and autonomous defense architectures in late 2026.