About
Aliri AI builds two layers of radiology intelligence. Aliri NLP turns free-text reports into structured, ontology-coded clinical data. Aliri+ sits on top as a conversational AI analyst over both clinical and operational data — ask in plain English, get cited answers in seconds.
Our Story
Aliri AI was founded to solve one specific problem: radiology departments produce vast volumes of clinical narrative every day, but most of that signal stays locked in unstructured text. Generic NLP tools weren't built for the language of radiology — the negation, laterality, temporal context, and precise anatomical specificity that clinical reports demand.
We built a purpose-designed NLP engine for radiology end-to-end. Our pipeline extracts clinical entities, resolves them against RadLex and SNOMED CT, and outputs structured data ready for cohort searching, follow-up management, critical findings detection, and quality assurance.
Today, Aliri AI processes millions of reports across healthcare organizations in North America, Australia, and beyond — embedded directly in radiologist workflows and integrated by analytics partners who bring our structured radiology data into their own products.
Our newest layer, Aliri+, takes the next step: let people talk to that data. Ask questions in plain English across clinical findings and operational BI — RVU, TAT, shifts, follow-ups — and get cited, grounded answers in seconds. No SQL, no dashboard-hunting. Available against a commercial LLM API or a local on-prem model when PHI safety is non-negotiable.
The Aliri AI Platform
Aliri AI is composed of two named components, designed to work together — the NLP pipeline that structures your data, and the AI analyst that lets you talk to it.
Layer 1 · Foundation
A purpose-built medical NLP pipeline for radiology. Ingests reports from RIS / PACS / HL7, extracts every clinical entity, and maps each one to RadLex and SNOMED CT — turning free text into structured, queryable, ontology-coded clinical data.
Layer 2 · Conversation
A conversational AI layer on top of Aliri NLP and your BI feeds. Ask questions in plain English across clinical findings and operational metrics (RVU, TAT, shifts) — and get cited, grounded answers in seconds. No SQL, no dashboard-hunting, no waiting on an analyst.
One platform, two layers. Aliri NLP builds the structured foundation; Aliri+ turns that foundation into a conversation — so the people who need answers can ask for them directly.
What Drives Us
Every decision at Aliri is guided by four core principles.
Every NLP model, every ontology mapping, every rule — built with input from radiologists and validated against real clinical data.
We don't invent proprietary taxonomies. We build on RadLex and SNOMED CT — the standards the industry already trusts.
Every feature ships with metrics. We measure entity accuracy, processing speed, follow-up completion rates — and we publish them.
Aliri runs on-premise or in your private cloud. Report data never leaves your network. HIPAA-compliant architecture from day one.
Integration Partner
PresiNET Systems integrates both layers of the Aliri AI platform — Aliri NLP for entity extraction and ontology mapping, and Aliri+ for conversational analytics — into its DImax analytics platform and Bialogics radiology intelligence products. Together they bring structured radiology data and plain-English AI Q&A directly into the clinical and operational workflows used by radiology departments across North America, Australia, and beyond.
Get in touch with our team to discuss how Aliri AI can work for your organization.
Contact Us