FREQUENTLYASKED

ClinixBoost is an AI-powered oncology software platform built specifically for community oncology practices — the practices that administer over 60% of U.S. chemotherapy but have historically been underserved by software designed for large hospital systems.

The platform has four main products: NCCN Query (real-time guideline answers at the point of care), ClinixBoost EHR (a full oncology-native electronic health record), the NCCN Compliance Checker (automated prior authorization), and Clinical Trial Matching (AI-powered trial eligibility across partner networks). Each can be used independently or together.

Our primary users are oncologists and clinical staff, practice administrators, and IPA/health plan teams involved in prior authorization review.

Both, depending on which product you use. The ClinixBoost EHR is a full platform replacement — scheduling, patient management, clinical documentation, dictation, messaging, billing, and a patient portal in a single system.

The NCCN Query and NCCN Compliance Checker can operate as supplemental tools alongside any existing EHR. NCCN Query is embeddable via a single API call, and the Compliance Checker accepts clinical notes by fax or email with no EHR integration required.

The NCCN Compliance Checker and NCCN Query cover 90+ cancer types with dedicated NCCN guideline coverage, including:

• Hematologic: AML, B-Cell Lymphoma, CLL, CML, Hodgkin, Myeloma, MDS, MPN, T-Cell, Waldenström’s
• Solid tumors: Breast, NSCLC, SCLC, Colon, Rectal, Pancreatic, HCC, Prostate, Bladder, Ovarian, Cervical, Melanoma, Head & Neck, Esophageal, Gastric, Renal Cell, Endometrial
• Pediatric: Pediatric ALL, B-Cell, CNS, Hodgkin, Neuroblastoma, Wilms Tumor
• Rare: Amyloidosis, Castleman, Histiocytic Neoplasms, Mastocytosis, Neuroendocrine, Occult Primary

Three meaningful differences:

• Built for community oncology, not adapted for it. Epic was built for hospital systems. Flatiron serves large practices with research infrastructure. ClinixBoost is designed from the ground up for the independent community oncology practice.
• AI is first-class, not bolted on. AI dictation, NCCN Query, and prior auth automation are core workflows — not integrations sold separately.
• Prior authorization is automated end-to-end. Most EHRs surface prior auth as a documentation task. ClinixBoost generates the compliance report, attaches NCCN citations, and delivers it via eFax without a staff member writing a letter.

Yes. ClinixBoost is built on a multi-tenant architecture from the ground up. Every record in the system is scoped to a practice at the database level. Practices are completely isolated — no data cross-contamination is possible by design.

Within a practice, multiple clinic locations are supported. The scheduling, patient management, and reporting systems are all clinic-scoped where appropriate.

ClinixBoost is at a late-prototype / pre-production stage with meaningful functional depth — full EHR workflow coverage and working AI dictation and compliance pipelines.

We are being honest about where we are: the platform is functionally rich but not yet in production deployment. We have a clear roadmap to production readiness, including mobile web support, frontend testing infrastructure, and BAA verification with PHI-handling vendors.

The dictation pipeline has four stages:

• Speech capture: A medical-domain speech recognition system converts speech to text using a model trained on clinical language, drug names, and oncology terminology.
• AI correction & enrichment: AI reviews the transcript and enriches it with patient context — demographics, current medications, active problem list, and recent encounters.
• Structured note generation: AI generates a formatted clinical document — SOAP note, consult note, or encounter summary — ready for provider review and signature.
• PDF export: The signed note is exported as a PDF and available for download or attachment to the patient record.

Speech recognition: A medical-domain speech recognition system selected for its accuracy on medical terminology, drug names, and oncology-specific language.

Language model (EHR): A primary language model selected for structured-extraction accuracy and a safety-focused development approach. Used for dictation correction, note generation, and NCCN Query.

Language models (Compliance Checker): A primary language model with a fallback model, orchestrated for resilience. All compliance checks run deterministically — same input, same output, every time.

We are model-agnostic by philosophy — we use the best tool for each clinical task.

NCCN Query grounds every answer in the actual NCCN guideline content — every answer cites the specific guideline section, version, and NCCN category (1, 2A, 2B, or 3) that supports it.

The system does not generate opinions or synthesize outside of the retrieved guideline text. If the guideline doesn’t address a query clearly, the system will say so rather than speculate.

Voice recordings are processed in real-time by the speech-recognition provider and are not stored by ClinixBoost beyond the active session. The transcript and generated note are stored in the ClinixBoost EHR database, encrypted at rest, scoped to your practice only.

Clinical notes processed through the Compliance Checker are transmitted to LLM provider APIs for analysis. We are in the process of verifying and documenting Business Associate Agreements (BAAs) with each PHI-handling vendor — a prerequisite we consider non-negotiable before production deployment.

Multilingual dictation is on our roadmap with an architecture designed and ready: explicit language selection, automatic language detection, and a real-time translator mode that accepts dictation in the provider’s preferred language and generates structured notes in English.

This feature is not yet in production. Current dictation is English-only. Contact us if multilingual support is a priority for your practice — it will directly inform our release timeline.

This reflects the core insight behind the Compliance Checker: the majority of prior auth denials occur not because the treatment is clinically wrong, but because the prior auth submission didn’t clearly tie the treatment to NCCN evidence in the format payers expect.

When every submission includes the specific NCCN guideline section, version, and category recommendation — along with patient-specific rationale extracted from clinical notes — payers have the information they need to approve on first review.

There are three submission paths:

• Web upload: Upload a PDF directly through the interface. Results appear in-browser and are delivered via email.
• eFax: Fax clinical notes to the ClinixBoost fax number. The pipeline monitors incoming faxes, splits multi-patient documents, processes each one, and returns a compliance report by eFax — with zero staff interaction.
• Email: Send clinical notes as email attachments. Same automated processing pipeline as fax.

Processing time from submission to delivered report is in minutes.

NCCN licensing for commercial use of guideline content is a prerequisite we take seriously and are actively addressing. Using NCCN guidelines in a commercial compliance product requires a formal agreement with NCCN — this must be verified and in place before any production deployment involving NCCN content.

Yes. The drug policy compliance workflow accepts any health plan’s drug policy PDF. The AI extracts three categories of criteria:

• General requirements: Baseline eligibility criteria that apply to any patient
• Specific indications: Condition-specific criteria (diagnosis, line of therapy, biomarker status)
• Exclusions: Contraindications or conditions that disqualify the patient

The system does not require pre-configured health plan templates — it reads and adapts to any well-structured drug policy PDF.

Yes — Clinical Trial Matching is a standalone product in the ClinixBoost suite (alongside NCCN Query, the EHR, and the Compliance Checker). Accepts three input methods: upload documents (PDF, JPG, PNG), voice dictation (medical-domain speech-to-text), or manual text entry.

The AI extracts structured eligibility data from unstructured notes — diagnosis, biomarkers, lab values, staging, treatment history, ECOG, comorbidities — then screens trials from partner networks.

When eligibility depends on missing information, the system generates targeted clarifying questions with AI-suggested answers (pre-filled from your notes with confidence indicators). Deep assessment returns criterion-level verdicts — Met, Unmet, or Uncertain — with full rationale, downloadable PDF reports, and one-click referral emails to trial coordinators.

See the Clinical Trial Matching page for the full workflow.

ClinixBoost is architected with HIPAA compliance as a first-order requirement — multi-tenant data isolation, role-based access control, encryption for messaging, encrypted storage, and audit logging are all built in.

However, full HIPAA compliance requires more than technical architecture. It also requires signed BAAs with all PHI-handling vendors, formal security policies, breach notification procedures, and a completed HIPAA risk assessment.

Multiple layers of protection are in place:

• Data isolation: Every patient record is scoped to a practice at the database level.
• Authentication: Industry-standard JWT with session limits and token refresh, plus role-based access control with granular permissions.
• Messaging encryption: Real-time messages are encrypted. Connections are authenticated at the protocol level.
• Transit encryption: All external communications use HTTPS/TLS.
• Storage encryption: Database and cloud storage are encrypted at rest.
• Audit logging: All data access and mutations are logged with user identity and timestamp.

Access is controlled by role-based access control with granular permissions, enforced at the API level. Roles include System Admin, Practice Admin, Provider, Nurse, Front Desk, Billing, and Patient (portal only).

BAA verification with all PHI-handling third-party services is a pre-production requirement we treat as non-negotiable. This verification is currently in progress and is not yet complete.

ClinixBoost does not use your clinical data to train AI models — ours or anyone else’s. Your patient records are stored in your practice’s isolated database partition and are not used for any purpose beyond providing the service to your practice.

Pricing is not yet publicly listed — we are in the early access phase and are structuring pricing in conversation with our first practice partners.

Our planned model is per-seat SaaS subscription for the EHR platform and NCCN Query, with per-report or volume pricing for the Compliance Checker as a standalone prior auth product.

Access is not self-serve at this stage. To request access, contact us at request.access@clinixboost.org with a brief description of your practice, your primary use case, and your expected timeline.

We review access requests and schedule a conversation to understand your workflow before onboarding. This ensures the platform is actually a fit before you invest time in an evaluation.

Yes. The NCCN Compliance Checker is a standalone product that requires no EHR integration. Clinical notes can be submitted by fax, email, or direct upload — no connection to your existing EHR system is needed.

This makes it particularly well-suited for IPAs, health plans, and billing services that want to automate prior authorization review without adopting a new EHR platform.

We offer guided demos tailored to your role and use case. Contact us to schedule one.

Free self-serve trials are not available at this stage. Given the clinical nature of the platform, a guided evaluation is more valuable than a time-limited login — we can focus the demo on the specific workflows that matter most to your practice.

Reach out at seek.help@clinixboost.org or use the Contact page to schedule a conversation.

Yes. NCCN Query exposes a single REST endpoint designed for EHR integration. Backend-to-backend authentication means your EMR can call it directly — no user login required.

The endpoint accepts a natural-language query, an optional domain parameter, and optional patient context. It returns a streaming response of cited answer tokens in real time. Drop it into any EMR workflow in an afternoon.

A native mobile app is on the roadmap with the shared platform foundation already in place — the same data model and APIs the web platform uses, with secure token storage on device.

Core navigation and authentication are scaffolded; functional UI is the next milestone. The web platform is mobile-accessible in the interim.

ClinixBoost runs on a modern, async-first cloud architecture: a Python backend, a server-rendered React frontend, role-based authentication, encrypted storage, and a horizontally scalable container deployment on a major cloud provider.

The AI layer uses multiple LLMs and a medical-domain speech-recognition system, orchestrated for resilience. We are model-agnostic by philosophy and use the best tool for each clinical task.

Not yet. FHIR/HL7 interoperability, ONC certification, e-prescribing, and insurance verification are on our medium-to-long-term roadmap, and we are transparent that their absence is currently one of the largest gaps relative to mature EHR competitors.

FHIR integration architecture has been researched and planned. Contact us if this is a blocker for your evaluation.