Airia vs Altilia (2026)

Stands at F, now on the vendor's own pages rather than a review aggregator. The claim is breadth at scale: connecting agents to thousands of data sources and enterprise systems already in use, delivered through a named Integration Framework of out-of-the-box connectors spanning SaaS platforms, databases and CRM systems, with Model Context Protocol support making the surface open-ended rather than fixed to a catalogue. The MCP direction matters and is graded here deliberately. Airia's MCP support is described as connecting the customer's systems to Airia, which is inbound tool consumption, so it credits integration breadth on this cell rather than extensibility on Ext, where the outbound direction would be needed and is absent. Keeping those apart is what let Ext drop to Partial on this record while this cell holds at Full. What is not documented, and worth noting for a comparison page, is the connector list itself: no catalogue enumerating which SaaS platforms and CRMs are covered appears on any surface reached, so thousands of data sources is the vendor's characterisation rather than a countable set. The grade rests on the framework and the MCP escape hatch being real, which they demonstrably are, rather than on the number.

F>P, and this is the closest call on the record, so the reasoning is recorded rather than left implicit. ONE FACT MOVED OFF THIS CELL AND IT WAS DOING MOST OF THE WORK. The July basis credited INGESTS ANY DOCUMENT OR DATA SOURCE here, listing PDFs, emails, handwritten notes, scanned images and legacy databases. That is ingestion for grounding and belongs on Know, where the Knowledge Graph carries Full. Reading a source is not integrating with a system the agent then acts in, and this record's real strength is the reading half. WHAT REMAINS IS A CATEGORY WITHOUT MEMBERS. API-first architecture and PRE-BUILT CONNECTORS are stated repeatedly, with agents EMBEDDED INTO YOUR ENTIRE TECH STACK, FROM LEGACY SYSTEMS TO MODERN CLOUD APPLICATIONS. Across two passes covering the home, platform, why-altilia, ai-agents and modules pages, not one connector is named. No CRM, ERP, ticketing or storage system appears, no catalogue exists, and no count is given. THE AXIS MEASURES BREADTH ACROSS CLASSES, and an unenumerated claim of breadth cannot evidence it. This is the same position ai-library was held at earlier today for the same reason, and consistency requires the same answer. WHAT DOES SUPPORT PARTIAL is real: the API-first architecture is a genuine integration route, agents perform actions rather than only answering, and the platform is distributed through the Microsoft commercial marketplace, which implies at least an Azure-side path. I RECORD THIS AS THE CELL MOST LIKELY TO BE UNDERSTATED ON THIS RECORD. A vendor with thirty-plus large enterprise clients in banking, insurance and public administration necessarily connects to core systems; the connector list simply is not published on the pages reached. A partner or integrations page would settle it and is the first check at lane close.

Stands at F, re-based off the vendor's own agent builder page rather than a vendor blog post about winning an award. The ROUTING ENGINE is the distinguishing mechanism: it distributes tasks to the most appropriate model, agent or interface, which makes routing a first-class platform service rather than something each builder wires by hand, and the vendor lists it among its six core capabilities as an operational reliability component rather than a convenience. Multi-agent composition is documented directly: the vendor describes designing COLLABORATIVE AGENT SYSTEMS that execute complex workflows while remaining visible, manageable and resilient as they grow in scope. That is the multi-agent half this axis looks for, and the emphasis on remaining manageable at scale is consistent with the governance-first positioning throughout the record. Both no-code and pro-code construction are supported in one environment so business and technical teams work on the same agents without breaking governance, and reusable logic patterns reduce policy drift across an estate. What is not documented anywhere is the control-flow primitive set, no branching, looping or conditional constructs are named, so I cannot say how expressive the builder is; the grade rests on multi-step orchestration and multi-agent composition being explicit, which they are.

Stands at F, and the orchestration primitive is named rather than described as drag-and-drop. VISUAL SCRIPTING IS THE MECHANISM: USE VISUAL SCRIPTING TO DESIGN AI AGENT WORKFLOWS, AUTOMATING AND SEQUENCING ACTIONS, CONNECTING DATA SOURCES WITH FINE-TUNED MODELS, CHAINS AND PROCESSING FUNCTIONS. Sequencing actions and composing chains against specific fine-tuned models is genuine multi-step construction, and binding a step to a particular model is a level of control most visual builders in this lane do not offer. SKILLS ARE THE VENDOR'S NAME FOR PROCESS ORCHESTRATION TASKS, built in the low-code interface and combined with agents; independent analyst coverage describes the agents-plus-skills combination as providing robotic process automation capabilities. The vendor's own framing of ONTOLOGY-DRIVEN WORKFLOWS ties the orchestration layer to the knowledge graph, so a process step can be defined against business entities rather than against raw data. THE MULTI-AGENT LAYER IS DOCUMENTED as governable agentic orchestration coordinating multiple agents across departments, and the secondary Multi-agent platform categorisation on this record is consistent with that. THE ARCHITECTURAL POINT WORTH CARRYING is the split the vendor makes between assistants and robots inside one orchestration model: a workflow can hand a step to an autonomous agent or route it through a person, which is why the oversight cell and this one describe the same visual surface from different angles. What is not documented is control-flow vocabulary. No branching, looping, conditional or parallel construct is named on any page reached, and no failure or retry behaviour is described. Sequencing and chaining are documented; expressiveness beyond that rests on the visual scripting claim.

Stands at P. Agents are documented as deploying into automated enterprise workflows across teams and systems, and the Routing Engine distributes tasks to the appropriate model, agent OR INTERFACE, which implies more than one surface exists, though the interfaces are never enumerated. What is absent from every surface reached is any named trigger class: no schedule or cron, no inbound webhook, no event subscription, and no messaging, email, chat or voice channel through which an agent is reached. For a platform positioned on operating agents in production across an enterprise, that is a conspicuous documentation gap rather than a plausible product gap; agents that run real work inside enterprise workflows are almost certainly started by something other than a person clicking, and the vendor simply does not say what. So this is Partial on retrievable evidence and it is the cell on this record I would most expect to move. Confidence is medium, the caveat being the same one that applies to moterra: Airia publishes no documentation site, so absence rests on marketing pages rather than a documentation index, which is materially weaker evidence. A deployment or integrations page enumerating triggers would settle it.

Stands at P. The July basis called channel coverage limited and that reading holds. THE EVENT CLASS IS DOCUMENTED THROUGH THE WORK ITSELF. Agents are triggered by document and data ingestion, with the platform ingesting from any source or format including legacy databases, emails, scans and forms. For a document automation platform that is the primary and correct entry route: work arrives because a document did. A SECOND ROUTE IS THE API. An API-first architecture with agents embedded into the customer's applications means an external system can invoke an agent, which is the inbound programmatic class and is credited on Ext as a surface but counts here as a route by which work reaches an agent. WHAT IS ABSENT IS THE REST OF THE SPREAD. No scheduled, cron or recurrence capability is named. No inbound webhook or external event subscription appears as a distinct capability. And no channel layer exists at all: no chat, email, messaging or embedded conversational surface through which a person reaches an agent is documented, beyond Altilia Insights as an in-platform assistant. THAT IS COHERENT FOR THE PRODUCT rather than a deficiency, and the grade should be read that way. This is back-office document and knowledge automation for regulated enterprises, not a conversational front end. Work arrives as documents and results go into systems; there is no customer sitting on WhatsApp waiting for a reply. A vendor with five channels and no knowledge graph would score better on this axis and be worse at the job Altilia does. Confidence medium: no triggers, scheduling or automation page was reached across two passes, and a platform doing overnight batch document processing for thirty enterprise clients almost certainly schedules, so this is likelier undocumented than absent.

Stands at P under the persistence line settled this session. What the vendor documents is CONNECTION rather than a maintained structure: agents connect to thousands of data sources and enterprise systems for real-time context, turning fragmented data into actionable intelligence. Reaching a live system when a question is asked is per-request assembly, which is precisely what holds tembo, lovable and moterra at Partial and what separates them from vellum's Document Index or bolt's synced design system. Nothing on any surface reached describes a knowledge base, an index, an embedding store, ingestion or retrieval tuning as a first-class object the customer maintains. Given how thoroughly this vendor documents its six security capabilities, the absence of any named knowledge object is meaningful rather than accidental: Airia's positioning is governing and routing agents that reach the customer's existing systems, not holding a corpus on their behalf. That is coherent and arguably the right architecture for a governance platform, since not copying the data is itself a security posture. But this axis measures grounding structure, and connection without persistence is Partial. Confidence is medium because the vendor publishes no documentation site and a knowledge or RAG capability could exist without appearing on marketing pages.

Stands at F and is the strongest cell on the record by a clear margin. It is also where several facts removed from other cells correctly land. THE KNOWLEDGE GRAPH IS THE PLATFORM'S CORE, not a retrieval add-on. The vendor states it AGGREGATES AND ORGANIZES ALL INFORMATION WITHIN A COMPANY, UNIFYING DATA FROM EVERY POSSIBLE SOURCE, STRUCTURED, UNSTRUCTURED, AND EVEN PHYSICAL DOCUMENTS, INTO A SINGLE COHERENT DATABASE. A graph is the strongest form of maintained retrieval structure there is, because it encodes relationships between entities rather than similarity between text chunks, and it is queryable by traversal rather than only by nearest neighbour. THE ARCHITECTURE IS HYBRID AND THAT IS THE DIFFERENTIATOR. Large and small language models are combined with the graph in a neuro-symbolic design, described as GRAPHRAG WITH TRAVERSAL REASONING. The symbolic half is what makes an answer traceable to a structure rather than to a probability. INGESTION BREADTH IS EXCEPTIONAL AND NOW SITS HERE rather than on Int: any source or format, including long complex documents, standardised forms, SCANNED IMAGES OR HANDWRITTEN NOTES, emails and legacy databases. Handwriting and scans matter for the regulated European buyers this targets, where the source of truth is frequently paper. CITATION IS DOCUMENTED, which few records in this lane manage: users interrogate the knowledge base in natural language and receive PRECISE, FACT-BASED ANSWERS WITH SOURCE ATTRIBUTION. Attribution is what makes a grounded answer checkable, and it is the property that carried over from Obs where it did not belong. RAG techniques retrieve and fine-tune LLMs and SLMs against enterprise knowledge, and the vendor reports above ninety-five percent accuracy on domain tasks, which is vendor-reported and recorded as such.

Stands at N, but the basis is rebuilt because the June one said a persistent cross-session memory COULD NOT BE VERIFIED, which is failed-pass language and manufactures absence rather than establishing it. Absence is now grounded in what the vendor does say. Airia's own account of how agents get context is real-time connection to enterprise data sources, not retention. Across the home page, the platform overview, the agent builder, deployment and FAQ pages, six security capabilities are named individually and no memory, state, session or persistence capability appears among them or anywhere else. For a vendor that documents its architecture this granularly, a first-class memory layer would be named if it existed. So N is the honest reading of retrievable evidence, and it is coherent with the positioning: a governance platform that deliberately does not copy customer data into its own store has a structural reason not to accumulate agent memory either. THE CAVEAT MATTERS THOUGH, and is the same one that applies to moterra and to Trig on this record: Airia publishes no documentation site, so this rests on marketing pages rather than a documentation index. Session persistence is a basic capability that frequently ships without being marketed, and this is among the likeliest cells on the record to be wrong.

Stands at P under the state limb of the 31 August ruling, and the interesting part is a distinction the July basis blurred. CONTINUOUS LEARNING IS NOT MEMORY, and this record is the clearest case for separating them. The vendor documents that THROUGH CONTINUOUS LEARNING AND HUMAN FEEDBACK, AGENTS BECOME INCREASINGLY ACCURATE, and that reviewed corrections feed improvement over time. Retaining corrections across runs is the accumulation limb of the ruling, and on the wording alone this would reach Full. BUT THE MECHANISM IS FINE-TUNING, NOT STATE. The platform page describes creating TRAINING DATASETS from real documents and using RAG techniques to RETRIEVE AND FINE-TUNE LLMS AND SLMS. Corrections are absorbed by updating model weights offline, not written to a store the agent reads at runtime. Those are genuinely different capabilities: a fine-tuned model is better at the task in general; a memory lets an agent recall what happened with this customer last month. Crediting weight updates as memory would let every platform with a feedback loop claim the axis. WHAT IS LEFT IS THIN. Altilia Insights is a conversational assistant, so session context exists within a conversation, which is the ruled definition of Partial. The Knowledge Graph persists and is addressable, but it is the grounding structure and is credited on Know; treating a corpus as agent memory would be the same double-count. NO CROSS-SESSION AGENT STATE IS DOCUMENTED: no memory store, no retention policy, no per-user context, no primitive by which an agent writes state in one run and reads it in another appears on any page reached across two passes. WORTH CARRYING FOR THE LANE: this is the second record this session where a learning loop presented as memory. The test that separates them is whether the artefact is read at inference time or absorbed into weights.

Stands at F, on the same reasoning that took lyzr to Full: automated policy enforcement at runtime is a shipped guardrail mechanism, and this axis credits guardrails as well as approval. Two of Airia's six named security capabilities are squarely on this axis. AGENT CONSTRAINTS is a policy engine controlling which data and which tools an agent may reach, launched by the vendor as a centralised governance layer, and RESPONSIBLE AI GUARDRAILS check outputs before they are acted on. Above both sits centralised policy inheritance: every agent inherits organisation-level policies governing model access, tool usage and deployment standards, and the vendor states policy controls are configured before an agent ever reaches production. That is oversight imposed by an administrator on other people's agents, which is stronger than a builder choosing to add a gate to their own, and it is the layer bolt lacked when I graded it earlier today. What is absent, and recorded rather than glossed, is the human half: no approval step, checkpoint, pause-for-confirmation or escalation queue appears on any surface reached. Compare agent.ai, graded today at Full on the opposite basis, shipping human confirmation actions and no policy engine. Airia has the enterprise half and not the personal one, which fits the buyer but means an agent that stays inside policy proceeds unsupervised.

Stands at F, and the mechanism is documented as a workflow stage rather than as a governance principle. THE REVIEW STEP IS EXPLICIT: REVIEW AND VALIDATE DATA CLASSIFICATIONS, EXTRACTIONS AND AI-GENERATED ANSWERS, with the vendor adding INTEGRATE THE FEEDBACK OF HUMAN EXPERTS TO ENHANCE ACCURACY AND CONTROL AT KEY STAGES. At key stages is the phrase that matters: the checkpoint sits at chosen points in the workflow rather than being a blanket setting, which is the same property that earned agentx its Full earlier today. THE ARCHITECTURAL DISTINCTION IS THE STRONGER EVIDENCE. Altilia ships agents in two declared forms, ASSISTANTS that support a human decision and ROBOTS that automate a process autonomously. A platform that makes the autonomy level a first-class choice at design time has built oversight into its model of what an agent is, rather than bolting an approval toggle onto an autonomous default. FOR THIS PRODUCT THE VALIDATION STATION IS THE RIGHT SHAPE. The work is document classification and extraction feeding downstream processes, so the consequential moment is accepting an extracted value, not calling an external tool. A human confirming a classification before it enters the knowledge graph is the approval gate that matters here, and it is where the vendor put it. THE FEEDBACK LOOP IS DOCUMENTED AS CLOSING: reviewed corrections feed continuous learning, so oversight improves the system rather than only catching individual errors. What is not documented is a policy layer: no confidence threshold routing low-certainty items to review, no action allowlist and no spend or scope limit appears on any page reached. Oversight here is placed by design rather than triggered by rule.

F>P on the section 7 hedge ladder, and the vendor contradicts itself across two of its own pages, so the resolution matters. The FAQ states that Airia supports BUILT-IN ALIGNMENT WITH GDPR, HIPAA, FERPA, COPPA, PCI DSS, SOC 2 (AICPA), ISO standards and the EU AI Act. Aligned with is named explicitly on the hedge ladder as Partial language, and it is doing a lot of work here: alignment with SOC 2 is a claim about how the product helps a customer meet a standard, not a claim that Airia has been audited against it. The deployment page separately says the managed SaaS is hosted on AWS, Azure and Google Cloud WITH SOC 2 TYPE II CERTIFIED OPERATIONS, which reads more like an attestation, but the sentence attaches the certification to the hosting arrangement and it is genuinely ambiguous whether Airia or its cloud providers hold it. A vendor with its own SOC 2 Type II report normally says so plainly and links it. So the attestation half is hedged or ambiguous on every surface reached, and F requires the real thing documented. THE CONTROL HALF IS EXCEPTIONAL and is why this sits at the top of Partial rather than lower: six named security capabilities, Security Posture Management, Agent Constraints, a Routing Engine, Agent Red Teaming, AI Discovery and Responsible AI Guardrails, plus role-based and granular access control and centrally inherited agent policies. A trust centre, a named audit firm or a linked report would move this to Full immediately.

F>P, and the vendor's own security paragraph is what decides it. The July basis read the certifications as held; the sentence says something weaker. THE EXACT WORDING: Altilia ensures robust data security through its Integrated Management System, ALIGNED WITH ISO/IEC 27001, ISO/IEC 27017, AND ISO/IEC 27018 STANDARDS. Aligned with is the second rung of the hedge ladder in section 7 and is named there explicitly as Partial. It says the management system was built to those standards; it does not say a certification body audited and issued against them. THE SELF-CONTRADICTION IS INSIDE ONE PARAGRAPH AND ACROSS TWO PAGES. The same sentence continues THESE CERTIFICATIONS VALIDATE, treating alignment as certification, and a separate page describes an ISO-CERTIFIED PLATFORM. Under section 7 the more precise and technical statement wins over the looser marketing one, and the precise statement is the one that names the management system and the three standard numbers. WHY THIS MATTERS RATHER THAN BEING PEDANTRY: for a vendor selling data sovereignty to European regulated buyers, whether an ISO certificate exists is the first procurement question, and the distinction between aligned and certified is exactly what a procurement team is checking. Recording it as certified when the vendor says aligned would put a claim on the index the vendor itself does not make. THE CONTROL HALF IS GENUINELY MET, which is why this is Partial and not lower: ACCESS CONTROLS, ROLE-BASED PERMISSIONS is named, alongside risk management, incident response, and continuous monitoring and audits. Under the conjunction bar, one half documented and one half asserted is Partial. Two facts moved off this cell as belonging elsewhere: data sovereignty and deployment location are graded on Dep, and customer IP ownership of models is a commercial term rather than a security control. No certificate number, issuing body, scope statement or trust page was reached. That is the single check that would move this to Full.

Stands at F and is the strongest cell on the record. AI DISCOVERY is the distinguishing capability and is rare enough to name: it inventories every agent in the organisation, including ones the vendor did not build, ones embedded inside licensed third-party software, and ones employees connected personally. The vendor's framing is that whether agents were purpose-built, commercially adopted or never approved, the same controls apply the moment Airia is deployed. Observing agents you did not deploy is a materially different proposition from logging your own runs, and only gumloop's Gumstack made a comparable claim this session. Around it sit unified analytics on usage, cost and impact, continuous real-time monitoring, and audit-ready reporting, so what happened, what it cost and whether it was compliant are all answerable. The item that clears the reporting-is-not-auditing line decisively is CONTINUOUS GOVERNANCE DOCUMENTATION AUTOMATICALLY GENERATED AND MAPPED to the EU AI Act, NIST AI RMF, SR 11-7, HIPAA, ISO 42001 and SOC 2. That is not a dashboard someone reads; it is evidence produced in the form an auditor asks for. Worth noting the asymmetry with Sec, which sits at Partial: Airia produces compliance evidence for its customers more convincingly than it evidences its own certification status.

F>P. The July basis cited AI ops management, monitoring hundreds of agents, debugging and observability, and explainable results. Reading the platform page against those, two of the four are the wrong subject and one belongs to another axis. WHAT THE MONITORING ACTUALLY WATCHES IS MODELS, NOT AGENT RUNS. The documented capability is MONITOR AND TRACK MODELS PERFORMANCE TO SPOT INEFFICIENCIES and EASILY MONITOR AND MAINTAIN UP-TO-DATE AI MODELS THROUGHOUT THEIR LIFECYCLE, with automated resource adjustments based on performance and workload. That is model operations, an accuracy-and-lifecycle discipline, and it is a real capability. It answers whether the model is performing well; it does not answer what a given agent did on a given document and why. EXPLAINABILITY BELONGS TO KNOW AND HAS BEEN MOVED. Verifiable answers with SOURCE ATTRIBUTION come from the Knowledge Graph and are a property of the grounding layer. Citing a source is not a record of execution, and counting it here would let the same architectural fact carry two axes. WHAT IS MISSING IS THE WHOLE OF THE AUDITING HALF. No run history, per-document processing record, agent action log, decision trace, retention period or export path appears on any page reached across two passes. THE GAP IS SHARPER THAN THE GRADE SUGGESTS FOR THIS BUYER. Altilia sells to regulated European enterprises processing licences, tax forms, tenders and public registries, and positions on deterministic auditable AI. Those are exactly the workloads where someone will later ask why a document was classified as it was. Model-level accuracy metrics do not answer that for an individual case. PARTIAL RATHER THAN NONE because model performance tracking is real, is customer-facing, and does tell an operator that something has degraded, which is more than several records in this batch offered.

P>F, and the June basis withheld Full for lack of an explicitly documented self-host or on-premises build, which is not what this axis requires. The rule names region selection, VPC, on-premises OR RUNNING INSIDE THE CUSTOMER'S OWN CLOUD as qualifying, and Airia publishes a dedicated Deployment Options page documenting exactly that. Airia can be deployed within the customer's own dedicated cloud environments across four providers: AWS, Microsoft Azure, Google Cloud Platform and Oracle Cloud Infrastructure. Four is more than any other record reviewed this session, most of which offer one. The GOVERNMENT CLOUD COVERAGE is the detail that settles it, with AWS GovCloud and Azure Government Cloud both named; those are FedRAMP-boundary environments that exist precisely for workloads that cannot sit in commercial regions, and a vendor does not list them unless the deployment is real. A fully managed SaaS option sits alongside on AWS, Azure and Google Cloud with a 99.9 percent uptime commitment and automated backup and disaster recovery, so the customer chooses between the vendor operating it and running it in their own account. That is the same managed-plus-customer-cloud shape that took gumloop to Full this session, and it is a stronger position than moterra's single-provider AWS deployment.

Stands at F and is among the better-evidenced Dep cells in the lane, because deployment control is the company's positioning rather than an enterprise-tier add-on. THREE MODES ARE DOCUMENTED CONSISTENTLY across the home, why-altilia and platform pages: SaaS, private cloud, and on-premises, described by the vendor as DEPLOYMENT SOVEREIGNTY. Repetition across pages matters here; several records this session rested a deployment claim on one sentence in one place. THE PART THAT MAKES IT SUBSTANTIVE RATHER THAN A HOSTING MENU is that the models travel with the deployment. Domain-tuned SLMs and LLMs are SERVED LOCALLY, so an on-premises customer is not running a thin client that ships documents to a hosted model for inference. Model inference is the leak in most agent platforms' residency story, and this one closes it by design. For a buyer processing tax forms and civil registries, that is the difference between a usable option and a nominal one. DATA SOVEREIGNTY IS STATED AS THE PURPOSE, with data kept within the customer's secure infrastructure and customers retaining ownership of data, models and assets. The company's whole market position is a sovereign European operating system for agentic AI competing against US hyperscalers, which is only coherent if the deployment claim is real. Per the 30 August ruling the sovereignty and residency properties are graded here alone and are deliberately not credited again on Sec, where the attestation now sits at Partial on its own evidence. What is not documented is region selection within the SaaS offering, installation requirements, or which components run customer-side in the on-premises mode.

Stands at F but the basis is rebuilt and one number has been dropped. The June basis cited more than two thousand five hundred prebuilt agent templates, sourced to Gartner Peer Insights, which is a review aggregator and not something section 7 treats as evidence. THE COUNT IS NOT REPEATED HERE because I could not confirm it on any vendor page, and an unverifiable figure of that size is exactly the kind of claim that ends up quoted on a comparison page. What the vendor does document supports Full under the wider bar, which asks whether packaged assets can be adopted ready-made by any route. The agent builder page describes leveraging TRUSTED AGENT TEMPLATES AND REUSABLE LOGIC PATTERNS to reduce policy drift and ensure consistent execution across the enterprise, and policy templates are referenced alongside them. That is vendor-supplied adoptable material, and the framing is unusual and coherent with the rest of the record: templates here exist to make governance consistent rather than to save build time, which is a different argument for the same feature. Confidence is medium because the template library itself was not seen. If a later pass finds a browsable catalogue, the count can be restored with a first-party citation; if it finds only a handful of patterns, this cell should be re-examined against the wider bar rather than assumed.

Stands at F under the 31 August bar, which asks whether the customer receives packaged assets ready to adopt. THE PREBUILT SET IS NAMED BY BUSINESS DOCUMENT RATHER THAN BY CAPABILITY, which is the stronger form. The vendor documents DEPLOY READY-TO-USE AGENTS FOR COMMON USE CASES SUCH AS EXTRACTING DATA FROM INVOICES, ORDERS, AND TRANSPORT DOCUMENTS. An invoice agent is a finished job; a toolkit is a starting point, and the difference is exactly what this axis measures. A SECOND, BROADER SET IS DOCUMENTED BY FUNCTION: specialised agents for document classification by semantic context, data extraction from unstructured sources, semantic search and question answering, summarisation into reports, and document generation for emails and operational content. Those are role-shaped rather than component-shaped. THE TWO-ROUTE STRUCTURE IS EXPLICIT AND IS WHY THIS CLEARS FULL RATHER THAN SITTING WHERE THE GENERATION-ONLY RECORDS LANDED THIS SESSION. The vendor states customers either deploy ready-to-use agents or use the IDE to create bespoke ones, so adoption of a packaged asset is a first-class path beside building, not a fallback. Kalcend dropped to None earlier today because generation from a prompt had replaced the catalogue entirely; here both exist. BLUEPRINTS FOR COMMON USE CASES sit alongside, and the vendor's positioning of a digital workforce implies role-shaped packaging rather than parts. What I did not reach is an enumerated gallery, so the size of the ready-to-use set rests on the named examples plus the vendor's description. Three named document types is thin as a catalogue but specific enough to be checkable, which is more than an unenumerated count would be.

Stands at F, now first-party rather than sourced to a review aggregator. The vendor states agents are built and run ACROSS ANY MODEL, FRAMEWORK AND DATA SOURCE, and a customer quote on the platform page puts it in the buyer's own words: the ability to use any model within a guided and controlled environment. Bring-your-own-model flexibility and multi-model orchestration are described consistently across pages. The Routing Engine is what lifts this above a model picker. It distributes tasks to the most appropriate model automatically, so model choice operates per task at runtime rather than being a setting fixed per agent, and it is positioned as an operational reliability capability, meaning routing also covers failover rather than only cost or quality. The governance framing is the part worth carrying to comparison pages. Because every agent inherits centralised policies governing MODEL ACCESS, an administrator decides which models the organisation may use while builders retain choice within that boundary. That is the same shape gumloop and zapier offer and it is the mature form of this axis: model flexibility and model control are usually presented as opposites, and the vendors that do both well are the ones enterprises can actually adopt.

Stands at F, and it is unusually well documented for this axis because model choice is the vendor's strategic position rather than a feature. THE BREADTH IS THE FULL RANGE: any commercial or open-source model, large and small, with the vendor's own framing being YOUR MODELS, YOUR CONTROL. Customers retain IP ownership of models trained on their data. THE SMALL-MODEL AND LOCAL-EXECUTION HALF IS THE DISTINCTIVE PART and separates this from a provider menu. Domain-tuned SLMs and LLMs are SERVED LOCALLY, with model selection driven by COST AND DATA TYPE. Choosing a small local model for a high-volume document class and a large hosted one for hard reasoning is a genuine economic lever, and running the small one inside the customer's own infrastructure is what makes it available to buyers who cannot send documents to a hyperscaler at all. FINE-TUNING COMPLETES IT: the platform retrieves and FINE-TUNES LLMS AND SLMS against enterprise knowledge, so the customer is not only choosing among models but shaping the one they choose. Few records in this lane document that. THE STRATEGIC CONTEXT IS WORTH CARRYING because it explains why this cell is strong rather than incidental. Altilia positions as a sovereign European alternative to US hyperscalers, and model portability is the technical substance behind that claim: a platform that could only call OpenAI could not make it. Per the standing convention, local model execution is credited here as customer model control and the deployment location it implies is graded on Dep, not counted twice. What is not documented is a named provider list, a model picker interface, or automatic routing between models at runtime, none of which this axis requires.

F>P against Mike's 30 August bar, which is explicit that Full requires the platform to be callable from outside through a documented API or SDK, WITH MCP AS CORROBORATION RATHER THAN THE BAR. Airia's documented extensibility does not clear that. What is documented first-party is MCP support, described as connecting the customer's systems to Airia, which is the INBOUND direction: Airia consuming external tools. That is integration breadth and is credited on Int. Nothing describes Airia publishing its own MCP server for external assistants to call, which is the direction that would corroborate Ext. The other documented item is pro-code development alongside the low-code builder, and extensibility for advanced teams. That is real, and it strongly implies an API exists, but implication is not documentation and no API reference, SDK, package or endpoint listing appears on any surface reached. The June basis asserted application programming interfaces without pointing at one. This is a Partial that should be easy to overturn: an enterprise orchestration platform almost certainly ships an API, and one documentation page would restore Full. But grading F on an inference would be exactly the failure this record already shows elsewhere, where two cells were graded from non-retrieval. Confidence medium because a developer portal may exist outside the marketing site.

Stands at F under Mike's 30 August bar, though it is the least verified of the Full cells on this record and the note should say so. WHAT SUPPORTS IT. The vendor states an API-FIRST ARCHITECTURE, and the direction is the Ext one: agents are CONNECTED DIRECTLY INTO YOUR OPERATIONS AND APPLICATIONS and EMBEDDED INTO YOUR ENTIRE TECH STACK, meaning the customer's own systems invoke Altilia agents rather than the reverse. API-first is a claim about how the platform is built rather than a feature bolted on, and every capability being reachable through the API is what the term means. SIMPLE SDKS ARE NAMED AS A FIRST-CLASS BUILD ROUTE, listed alongside natural language and visual scripting as the three ways a customer creates agents. A vendor offering an SDK as one of three peer entry points is describing a developer surface, not an afterthought. CONFIDENCE IS MEDIUM AND THE GAP IS THE SAME ONE I NAMED ON AGENTX TODAY: no API reference, endpoint list, authentication documentation or SDK package was reached across two passes. The API is documented as an architectural property and a deployment route rather than through a developer surface a reader can inspect. I have graded consistently with agentx, where one-click-to-API carried Full on the same basis, rather than applying a stricter test here. No MCP server was found, which under the ruling does not withhold the grade. THE ASYMMETRY WORTH RECORDING: this vendor documents its knowledge architecture in unusual depth and its developer surface barely at all, which is consistent with selling to business buyers through enterprise engagements rather than to developers. A developer portal is the natural check at lane close, alongside the connector catalogue for Int.

Stands at P under the 31 August Eval bar, and the reasoning turns on what kind of testing is being done. Two capabilities are documented and both are real. A PROTOTYPING STUDIO provides secure controlled environments to experiment, iterate and deploy agents, which is a development loop. AGENT RED TEAMING simulates adversarial attacks to stress-test agent behaviour, which is genuine automated testing and more than most vendors in this lane ship. But red teaming answers can this agent be made to misbehave, not is this agent's output any good, and neither capability produces a score, pass rate, judge verdict or measured comparison across versions that a customer reads. Under the bar that is a quality gate without a readable comparable result, which is Partial. It is the same distinction that put moterra's compliance scanning and replit's self-testing at Partial today. Worth stating plainly because it is a real asymmetry on this record: Airia measures SAFETY thoroughly and QUALITY not at all. For a platform sold on governance that is a defensible priority, and a buyer should understand they are getting adversarial assurance rather than accuracy assurance. Confidence medium because a documented scoring or evaluation surface could exist outside the marketing pages reached.

Stands at P under the 31 August Eval bar. The July basis was accurate and the refinement here is about which half is present. THE MEASUREMENT SURFACE IS REAL BUT AIMED AT MODELS. The vendor documents MONITOR AND TRACK MODELS PERFORMANCE TO SPOT INEFFICIENCIES, ENABLE CONTINUOUS LEARNING AND IMPROVEMENT OVER TIME, alongside human review and validation of classifications, extractions and generated answers. Reviewed corrections both catch individual errors and feed the training loop, so quality is measured and acted on. WHY IT DOES NOT REACH FULL. The ruled bar asks for a result the customer can read AND COMPARE about the agent's behaviour on their own work. Nothing documents a test set held aside, expected outputs recorded, a scored run, or a comparison between one version of an agent and another. Model performance tracking tells an operator that accuracy has moved; it does not let them establish that a specific change caused it, which is the question a regression harness answers. THE VENDOR-REPORTED ACCURACY FIGURE IS RECORDED AND NOT CREDITED. Above ninety-five percent on domain-specific tasks is a vendor benchmark, and under the standing convention vendor self-benchmarking is Partial-class evidence at best, the same handling applied to codebuff. THE STRAIN WORTH NAMING for a platform of this kind: the review station generates exactly the material a golden dataset needs, since every human validation is an expected output recorded against a real document. Turning that into a scored regression suite is a small step from what exists, and its absence is more likely undocumented than unbuilt. Confidence medium; no evaluation, testing or debugging page was reached across two passes, and the debugging dashboard referenced in the July note was not found on any page this pass.

Stands at N, with the basis rebuilt because the June one used could-not-be-verified language, which is a failed pass rather than a finding. Absence is now read from what the vendor documents about how agents act. Every action path Airia describes runs through a programmatic interface: an Integration Framework of connectors to SaaS platforms, databases and CRM systems, Model Context Protocol support, and connection to thousands of enterprise data sources. Agent Constraints govern which DATA AND TOOLS an agent may reach, which is a tool-permission model rather than a screen-control one. Nothing describes a browser, a rendered interface, navigation, form filling or session control. The positioning makes this coherent rather than surprising. Airia's proposition is bringing agents inside a governance perimeter where every action is policy-checked and inventoried; driving software through its user interface is the opposite of that, because it produces actions that cannot be constrained by a tool policy or attributed cleanly in an audit trail. A governance-first platform has good reasons not to offer computer use. Confidence is medium rather than high for the standing reason on this record: no documentation site exists, so absence rests on marketing pages rather than a complete index.

Stands at N, and confidence rises to high because a genuine near-miss was found and refused rather than nothing being found. THE NEAR-MISS IS THE STRONGEST ON THIS RECORD AND WOULD CATCH A GRADER SKIMMING. Altilia calls its autonomous agents ROBOTS, distinguishing them from assistants, and independent analyst coverage states that the combination of agents and skills PROVIDES ROBOTIC PROCESS AUTOMATION CAPABILITIES. A vendor whose own vocabulary is robots and RPA, in a market where RPA historically means screen automation, reads at first glance like the positive case for this axis. IT IS NOT, AND THE DISTINCTION IS THE ONE THAT MATTERS FOR THIS LANE. Altilia's robots are autonomous agents that classify, extract, summarise and generate against documents and a knowledge graph, reaching systems through an API-first architecture and connectors. RPA capabilities here means the outcome replaces the manual work RPA was bought for, not that the mechanism is screen driving. No browser control, navigation, form filling, screen interaction, visual grounding or desktop automation appears on any page reached across two passes. COMPUTER VISION IS THE SECOND NEAR-MISS AND IS ALSO REFUSED. The platform uses computer vision and document analysis and recognition to read scanned images and handwritten notes. Reading a document image is intake, credited on Know, and is the opposite of operating an interface: the page is being interpreted, not driven. THE ABSENCE IS STRUCTURAL. A platform whose entire premise is turning unstructured sources into a queryable knowledge graph reaches data by ingestion and API, which makes screen automation unnecessary. WORTH CARRYING TO ENTERPRISE OPERATIONS, where IDP and RPA-adjacent vendors will recur: robots in the vendor's vocabulary and RPA in an analyst's are not evidence of computer use, and the test is whether an interface is being driven.