DNAI · RheumaAI · Clinical AI

Clinical intelligence for the rheumatologist who already knows
what they're doing.

Built from Dr. Erick Zamora's actual diagnostic protocols. Rheumatology-specific, not generic medical AI. Differential support, treatment recall, and clinical reasoning — grounded in real practice.

Request clinical access Read the brief
Proof of concept · In clinical evaluation · Dr. Erick Zamora
RheumaAI · Asistente Clínico · Dr. Zamora En uso

Caso clínico

María R. · 34 años · F Derivada por artralgia bilateral persistente 6 meses. Sin diagnóstico previo.

Presentación

Artralgia simétrica MCF y IFP. Rigidez matutina 90 min. Fatiga crónica. Fotíca moderada.

Laboratorio

ANA

+ (1:320)

anti-dsDNA

+ elevado

FR

Negativo

C3 / C4

Disminuidos

Diagnóstico diferencial

Alta probabilidad Lupus Eritematoso Sistémico Criterios SLICC: ANA+, anti-dsDNA+, C3/C4 bajos, artritis, fotosensibilidad. 4/11 confirmados.
Considerar Síndrome de Sjögren Solapamiento posible. Solicitar anti-Ro/La para descartar.
Bajo sospecha Artritis Reumatoide FR negativo, patrón sistémico sugiere LES.

Protocolo sugerido

Completar SLICC: solicitar anti-Sm, anti-Ro/La

SLEDAI-2K basal antes de iniciar tratamiento

HCQ 200-400 mg/d si dx LES confirmado

Nefroólogo si proteinuria en próxima búsqueda

"RheumaAI doesn't replace clinical judgment. It runs alongside it. Every output reflects what I actually do in practice, not what a textbook says."

Dr. Erick Zamora - Rheumatologist · Primary operator

1 of 7 products
in the DNAI
proof stack

Rheumatology is the hardest differential in medicine.

The core problem

Overlapping autoimmune presentations leave no room for generic AI guidance.

Lupus presents as RA. Sjögren overlaps with both. ANCA vasculitis mimics infection. A clinical AI that doesn't know Dr. Erick's specific patient population and diagnostic thresholds is not useful — it's noise. RheumaAI is built from real clinical decisions, not general medical literature.

Differential burden

Complex differentials cost consultation time.

Mapping ANA patterns, complement levels, and clinical criteria in real time slows the consultation. RheumaAI runs the differential while the physician examines.

Protocol recall

Drug monitoring and biologics protocols are high-stakes.

Methotrexate dosing, biologic switching criteria, infection screening thresholds. RheumaAI surfaces the right protocol at query time without a manual lookup.

How RheumaAI works.

01

Seeded from Dr. Erick's clinical knowledge.

RheumaAI doesn't use generic medical AI training data as its primary source. It is seeded with Dr. Erick Zamora's actual diagnostic protocols, his patient population characteristics, his threshold decisions for common presentations. The AI reasons the way the physician reasons.

02

Query with a real patient presentation.

Enter the clinical case — lab results, symptoms, timeline, prior history. RheumaAI maps the differential, surfaces relevant criteria (SLICC, ACR/EULAR, CARRA), and flags what's missing from the workup. All in the context of how Dr. Erick's practice actually operates.

03

Structured clinical output — not a chat response.

Differential ranked by probability with clinical rationale. Protocol steps in the right order. Flags for high-risk patterns. The output is structured for clinical use, not for a conversation. It fits into the consultation without disrupting it.

Clinical capabilities

What RheumaAI does today.

Current capabilities

  • Differential diagnosis support for common rheumatology presentations (LES, AR, Sjögren, vasculitis, espondilitis)
  • ANA pattern interpretation and complement consumption reasoning
  • SLICC, ACR/EULAR, and SLEDAI-2K criteria mapping in real time
  • Treatment protocol recall: DMARDs, biologics, dosing, switching criteria
  • Drug interaction and monitoring schedule surfacing
  • Clinical documentation assistance in Spanish and English
  • Structured case summaries for referrals and records

Access status

Proof of concept in active clinical evaluation.

RheumaAI is currently in clinical evaluation by Dr. Erick Zamora. Access is available to rheumatologists interested in evaluating the DNAI operator model for clinical AI in their practice.

Request clinical access Read the DNAI brief first

Part of the DNAI proof stack

RheumaAI LES AI RheumaScore ORVS Beach Science / clawRxiv STORM BiobadamexAI

Common questions.

Is this a general medical AI?

No. RheumaAI is built specifically from Dr. Erick Zamora's clinical knowledge base. It does not use a general medical AI layer as its primary reasoning source. It is trained on one rheumatologist's actual diagnostic decisions, protocol preferences, and patient population. This makes it specific and useful for rheumatology, and not useful for anything else.

What is the difference between RheumaAI and a clinical decision support tool?

Clinical decision support tools use general medical guidelines. RheumaAI uses Dr. Erick's specific clinical reasoning. It knows which ANA patterns he considers significant in his patient population, which biologics he prefers for specific presentations, and how he sequences workup steps. That specificity is the difference between useful and generic.

Is it approved for clinical use?

RheumaAI is a proof of concept in active clinical evaluation. It is not a regulated medical device. It is being used as a clinical reasoning support tool by Dr. Erick Zamora in his practice. Rheumatologists interested in evaluating the model for their own practice can request access through DNAI.

What does it output?

Ranked differential diagnoses with clinical rationale, specific diagnostic criteria mapping (SLICC, ACR/EULAR, SLEDAI-2K), suggested workup steps in the right order, treatment protocol suggestions, drug interaction flags, and structured case summaries. Outputs are in Spanish or English depending on how the query is entered.