V20 · Next-generation release

The AI-first Survey Operating System

Ask the right questions. Collect anywhere. Understand instantly. Act responsibly. Even when the internet isn’t there.

Built for humanitarian, M&E, research, and enterprise teams. Nonprofit tier includes 200 free AI credits / month.

One platform, five pillars

V20 reorganises everything Survona does into a clear product narrative you can sell, ship, and operate.

01 — Create

AI copilot builds your survey with you

Describe the outcome you need. Survona drafts the questionnaire, flags bias, translates into 40+ languages, and generates a live preview you can edit by talking to it.

Drag-and-drop builder v2 with logic debugger
AI authoring copilot (GPT-5.4 + Claude Sonnet 4.5)
Agentery 8-stage pipeline — 14 built-in app types
Live UI edits via Claude Opus 4.7 + Claude Design

Explore Create

02 — Collect

Anywhere your respondents are — even offline

Web, PWA, native iOS/Android, QR, SMS, WhatsApp, email, AI phone calls, kiosks, enumerator mode. Data syncs the moment a connection returns.

Progressive Web App — works fully offline
Native iOS + Android (React Native + Expo)
AI phone-call surveys in 40+ languages
Delta sync with CRDT conflict-free merges

Explore Collect

03 — Understand

Research-grade analytics + conversational insight

Categorai classification, themes, chat-with-data, misinformation detection, persona lab, benchmarks, and cross-tabs. Every answer cites its source rows.

Categorai v2 — 6-category standard + custom taxonomies
Misinformation Virality Detector (4 categories)
Chat-with-data: quick / standard / deep modes
F1 / MCC / precision / recall / confusion matrix

Explore Understand

04 — Act

Turn insight into action automatically

CARM cases, SLA timers, escalations, follow-up surveys, alerts, and workflow automations. Editorio publishes outputs to 22 channels in the right tone for the right audience.

CARM with SLA timers + escalations
Editorio — 47 content types, 22 output channels
5 report templates × 7 export formats (PDF, DOCX, XLSX…)
Public REST API + webhooks + SDKs

Explore Act

05 — Control

Enterprise-grade governance and trust

SSO, SCIM, audit logs, RBAC (9 roles), customer-managed keys, data residency, AI evaluation dashboards, and a nonprofit tier that stays free.

9 roles at org + workspace level
Central credit pools across country offices
SOC 2 Type II + EU AI Act conformity
Nonprofit tier — 200 free AI credits / month

Explore Control

Authoritative AI stack (April 2026)

The right model for each job

We keep your proven stack — GPT-5.4, Whisper-3, embeddings — and add Claude Opus 4.7 + Claude Design strictly for live UI/UX generation.

Workload	Model	Notes
UI/UX live builder & design	Claude Opus 4.7 + Claude Design	Launched 16–17 Apr 2026
Survey authoring copilot	GPT-5.4 (primary) / Claude Sonnet 4.5 (fallback)	Streaming diffable patches
Classification & misinformation	GPT-5-mini (batch) + GPT-5.4 (quality tier)	Batch 20 / call, parallel
Chat-with-data & deep analysis	GPT-5-thinking	Citations for every answer
Voice phone surveys	GPT-5 Realtime Voice + Whisper-3	40+ languages
Embeddings	text-embedding-3-large	pgvector + hybrid retrieval

Works where the internet doesn’t

PWA offline shell · native iOS/Android · AI phone-call interviewer · low-bandwidth analyst mode · signed .survonaexport packages distributable by USB or LAN.

See the offline stack

V20 ships every commitment in the Mercy Corps ECHO 2026 technical brief — Agentery, Categorai, Misinformation Detector, Editorio, automated reporting, marketplace, 9 roles, credit pools, and pilot pricing from $5 / mo.