Architecture
One platform, from your data to finished work.
Ingestion, the knowledge layer, agents and workflows, card apps and models, with security, governance, audit, monitoring and cost control under every step, so your data and intellectual property stay yours end to end.
The Inovagio platform
One platformAgent HarnessAgent harness
Inovagio CLIInovagio DesktopWeb ChatInovagio SDK
OpsisExperience layer
FolioMorphe
Virtual EmployeesVirtual Employee layer
NexusErgon
SynesisKnowledge layer
PerceptionMNEMOS
OriginModel layer
Inference ServerForge
PhylaxTrust layer
AuthInovagio SecretsGuard
OperationsOperations
PanoptesArchon
Platform servicesPlatform services
ServerAPI ManagementConnect
How it works together
From your data to finished work, in one system.
Ingest
Perception reads your sources
Documents, images, spreadsheets and the systems you connect are read, structured and stored with their source.
Know
Synesis learns the business
MNEMOS holds the data; agents interview your people and write the rules your experts approve.
Act
Virtual Employees do the work
Company goals are broken into tasks that agents run as workflows and jobs, talking to each other and asking a person when policy says so.
See
Opsis shows the result
Card apps, documents, spreadsheets, charts and 3D, every value linked to the record it came from.
Anywhere
The Agent Harness puts it in reach
The same agents in the terminal, on the desktop, in web chat or inside your own app, with one admin.
How work flows through Inovagio.
From the moment work arrives to the finished result, with every step aligned to your goals and recorded. Click the diagram to enlarge it.
Trust and operations run through every step.
Nothing reaches an agent, a model or a record without passing the same identity, policy and audit, and nothing runs out of sight.
- Trust layerPhylaxAuth for every person, service and agent; Inovagio Secrets for secrets and keys; Guard for safe inputs and outputs; policy and a tamper-evident record.
- ObservabilityPanoptesEvery product’s logs, traces and metrics in one place, with security, network, token and cost views.
- Agent controlArchonSee every running agent, goal and workflow, stop or start any of them, and pull a kill switch.
- Connects it allPlatform servicesThe Server, API Management and Connect serve the console and the APIs, govern every call and reach your systems.
Built for regulated industries from the foundation up.
Banks, insurers, health systems, law firms and agencies cannot send their records, processes or know-how to a system they do not control, and must show how every decision was made. So these five commitments are in every product, not added at the end.
Security
Every request is bound to the signed-in identity in code; secrets live in Inovagio Secrets; prompts, records and models stay in your environment.
Governance
Policy is checked before and after every protected action, and anything outside it waits for a named person.
Auditability
Each step records what was read, which rule applied, who approved it and what it cost, in a tamper-evident trail you can replay.
Observability
Every product sends its logs, traces and metrics to Panoptes, and Archon can see, stop or start any agent.
Cost per task
Routine work moves to small models you own, caches cut repeated work, and the cost of every task is in front of you.
Seven things that need the whole platform.
Rule Capture
Synesis
Interviews and records become rules your experts approve, each linked to where it came from.
Goal Alignment
Virtual Employees
Every task traces up to the company goal it serves.
Shadow Mode
Virtual Employees
AI drafts while your people decide, until accuracy holds on your own cases.
Safety Floor
Agent Harness
Dangerous actions are blocked in every mode, whatever an agent is allowed to do.
Kill Switch
Archon
Stop one Virtual Employee, a department or everything at once, and start them again.
Evidence Ledger
Phylax
A tamper-evident record of every step, ready to replay for an auditor.
Cost Ledger
Panoptes
The cost of every task, agent and workflow, in front of you.
One platform, built as one system.
Every piece a team would otherwise license, connect and secure separately is part of one platform, with one identity, one policy and one record.
| What teams stitch together today | With Inovagio | |
|---|---|---|
| Data | Separate relational, document, vector and search databases, a cache and a message broker | MNEMOS: one audited engine for all of them |
| Knowledge | Documents an assistant searches and guesses at | Synesis: rules your experts approved, with their sources |
| Agents and workflows | An assistant per team and an agent framework per project | Virtual Employees and the Agent Harness: one runtime, every face |
| Models | Hosted model APIs that see your data | Origin: small models you own on your hardware; large models only when you choose |
| Identity and APIs | A separate identity provider, API gateway, web server and secrets manager | Auth, API Management, the Server and Inovagio Secrets, under one configuration |
| Safety and governance | A safety add-on and a governance tool that reads logs after the fact | Phylax checks every action before and after it runs and keeps the evidence, and Guard screens what goes in and out |
| Operations | Separate monitoring, tracing and cost tools | Panoptes and Archon, from one record |
From a laptop to your data center, with the same software.
- On a laptop or workstation: the CLI, the Desktop app, the inference server and MNEMOS in one app.
- On your own servers or private cloud, for a team or the whole organization.
- Hosted for you, on Inovagio’s own servers and GPUs, shared or dedicated.
- Open small models on your own hardware, large models on GPU servers, and a frontier model only when you choose.
- Your identity provider, your monitoring and your keys, wherever it runs.
Articles and guides on the architecture.
Guide · Context engineeringContext engineering for enterprise AI agentsContext engineering decides what an AI agent sees at each step of a task. The research on why long contexts fail, the techniques that keep context small, how agent memory fits in, and the controls that stop context from becoming an attack path. · updated November 18, 2026 · 18 min read
Guide · Knowledge graphHow AI agents use your knowledge graphAn AI agent uses your knowledge graph in five steps. Approved terms become concepts, the ontology types the graph, retrieval finds connected records and their evidence, your rules check the result, and failures go to a person. How each step works, and what the GraphRAG research shows. · 17 min read
Guide · KnowledgeKnowledge layer: questions and answersStraight answers to the questions buyers and architects ask about the knowledge layer for AI agents, from ontologies, knowledge graphs and semantic layers to context engineering, GraphRAG, semantic caching, small models and who owns it all. · updated November 18, 2026 · 12 min read
Guide · OntologyOntology for AI agents: how to build one in five stagesHow to build an ontology AI agents can use, in five gated stages from a controlled vocabulary to a knowledge graph. The W3C standards for each stage, competency questions, Turtle and SPARQL for a worked example, entity resolution, versioning and "as of" queries. · updated November 18, 2026 · 21 min read
Guide · Data qualityPreparing data for a knowledge layer: a step-by-step guideHow to get data ready for a knowledge layer and the AI agents that read it, in the order that works. Scope by the first queue, profile against named quality dimensions, name owners, resolve entities reversibly, clean documents, record lineage, map to the ontology and put a contract on every source. · updated November 18, 2026 · 17 min read
Guide · Semantic layerSemantic layer for AI agents: governed metrics for accurate answersA semantic layer defines each business metric once, with its joins and its owner, so an AI agent asks for "credit note rate" instead of writing its own SQL. What the 2023 to 2026 benchmarks show, a worked metric definition, the open Ossie standard and where MCP fits. · updated November 20, 2026 · 14 min read
See the platform run on one of your workflows.
A 30-minute walkthrough of one task from your data to the finished result: the knowledge, the agents, the approval step, the record and the cost.