Frequently Asked Questions

Verixa is an AI-governed quality platform for pharma and biotech. Phase 1 is GMP Core — six workflows (Document Control, Deviation, Root Cause Analysis, CAPA, Findings, Inspection Readiness) — with MIRA, our quality copilot, as an advisory, human-in-the-loop assistant. Verixa is verified under our SDLC and is in design-partner validation.

Navira Quality Systems Pvt Ltd is the operating entity and owns the Verixa platform IP. Verixa is the platform. Sense7ai is the implementation partner that built the platform under contract and retains its general engineering IP per an implementation-partner agreement.

Sense7ai is the implementation partner; access to a customer’s tenant data is limited to defined support purposes under the relevant agreements and is covered in the DPA. Tenants are isolated, and customer data is not used to train shared models.

Phase 1 (now, in design-partner validation): GMP Core — Document Control, Deviation, RCA, CAPA, Findings, Inspection Readiness. Phase 2 (2027): GMP depth — additional cohort modules such as Change Control, OOS/OOT, Audit Management, Supplier Quality, APQR, Batch Records, Training, Mock Inspection. Phase 3 (2028+): GCP, GVP, and GDP, subject to customer pull.

The founding-partner engagement is a 90 day controlled GMP evidence sprint on one workflow, preceded by a 10-business-day Phase 0 Evidence Mapping. Timelines beyond the founding cohort are scoped per organisation.

Not yet. Those map to GCP, GVP, and GDP, which are Phase 3 (2028+) on the roadmap. Today Verixa is focused on GMP. You can register interest for a Phase 3 domain through the roadmap pages.

Verixa provides 21 CFR Part 11-aligned controls (e-signature, audit trail, access controls). Compliance itself is a customer determination based on intended-use validation, not a vendor claim — verification of the e-signature substrate is being completed in the design-partner phase.

No vendor can validate a system for your intended use — that is customer-owned. Verixa is verified under our SDLC and is currently in pre-validation, in design-partner validation. We provide the evidence and templates (the G-7 pack) that accelerate your validation; the customer validates intended use.

No. MIRA is advisory and human-in-the-loop. MIRA never approves, closes, signs, disposes, or releases a GxP record, and never makes critical GMP decisions (batch release, OOS classification, CAPA disposition). A qualified human accepts, rejects, or modifies every suggestion before it has any GxP impact.

Verixa treats EU GMP Annex 22 (Draft 2025) as an internal-control reference. Consistent with the draft, generative/probabilistic AI is prohibited from making — or being the sole path to — critical GMP decisions in Verixa. Annex 22 is a draft; we track it as it moves toward final.

Verixa makes no claim of EU AI Act compliance. Obligations under Regulation (EU) 2024/1689 are role-specific (provider vs deployer) and obligation-specific (transparency, human oversight, dependence on general-purpose models), and are phasing in through 2026–2027. MIRA is advisory with human-in-the-loop.

Verixa captures AI-assisted decisions as first-class audit-trail objects — model version, prompt hash, retrieved evidence, the human accept/reject/modify decision, and e-signature — an architecture designed to be inspection-defensible. Whether your records are audit-ready in a given inspection depends on your own intended-use validation; that determination is customer-owned.

Verixa’s validation accelerator (the G-7 evidence pack and templates) is designed to reduce the effort your team spends validating the system for its intended use. Industry benchmarks (CloudTheApp) put typical Year-1 CSV at $50K–$150K; the accelerator targets a 30–60% reduction of that effort. Actual customer outcomes will be published after design-partner validation. The customer still owns intended-use validation.

US tenants are hosted on AWS US-East; India tenants on AWS Mumbai, with DPDP-aware data handling. Tenants are isolated from one another.

Verixa uses Anthropic and Azure OpenAI as AI providers, hosted on AWS infrastructure; these are listed as subprocessors in the DPA. Customer data is not used to train shared models, and every inference logs the model name and version, prompt version, retrieval-corpus version, and guardrail version.

You keep a 30-day clean-exit export. The deliverables from the engagement — workflow map, evidence pack, configured templates — are yours to keep.

Verixa is built API-first, and integration with your existing systems (including single sign-on) is scoped with each design partner during Phase 1. Tell us your stack in the application and we’ll map the integration path.

India’s revised Schedule M raises GMP expectations for domestic manufacturers and CDMOs. Verixa Phase 1 GMP Core targets export-quality readiness for India CDMOs facing dual CDSCO and global-client scrutiny, with INR-denominated pricing, AWS Mumbai data residency, and DPDP-aware handling.

The Founding Partner Program is limited to 2 partners by 31 December 2026. Apply through the Design Partner page (a short qualifying application); a 10-business-day Phase 0 Evidence Mapping precedes the 90 day founding pilot.

Year-1 license bands: Emerging / Clinical Biotech US $15K–$60K (₹10–30L); Mid-Market Pharma/CDMO US $110K–$350K (₹30–150L); Enterprise US $350K–$1.5M+ (₹1.5–8 Cr) — all + applicable GST. AI is bundled — no token meter, no overage. The paid pilot ($11K / ₹9L, 100% credited) applies to Mid-Market & Enterprise; Emerging/Clinical Biotech begins with Phase 0 Evidence Mapping ($3.5K / ₹3L). Bands are post-design-partner-validation. See the Pricing page for detail.

Pricing is site-anchored, with unlimited users within a band. Multi-site organisations are quoted across sites — talk to us through the Pricing or Contact page for a scoped figure.