RESVANTA, INC. (DE, USA) · RESVANTA TECHNOLOGIES INC. (BC, CAN)
RESVANTA GROUP/RESEARCH ENGINE (RRE)/AI SURVEY PROGRAMMING (AISP)
TECHNICAL CAPABILITY DOSSIER // CIPO PATENT PENDING CA 3,329,739

AI Survey Programming from OpenXML (`.docx`) in Under 15 Minutes

The Resvanta Research Engine (`survey.resvanta.com`) eliminates the 24–48 hour manual survey scripting bottleneck without sacrificing routing accuracy. Powered by the Autonomous Intelligence Survey Programmer (AISP), researchers upload a standard OpenXML (`.docx`) questionnaire, resolve ambiguous logic inside an interactive Clarification Workspace, verify every skip path with synthetic respondents, and launch a live fieldwork link in under 15 minutes.

01 // THE SURVEY PROGRAMMING BOTTLENECK

Why Both Manual Scripting Shops and “Blind-Guess” AI Tools Fail Researchers

Market research questionnaires are authored in OpenXML (`.docx`) by insights directors, economists, and brand strategists—not software engineers. Historically, turning that document into a live survey with accurate quotas, piping, and termination logic forced teams to choose between two flawed options:

Option A // Manual Programming Bureaus
24–48 Hours per Draft

Outsourcing scripting to manual survey programming agencies or legacy enterprise service desks costs US$3,500–$8,500 per study and requires 1–3 business days of back-and-forth email revisions before link testing can even begin.

Option B // 5-Minute “Blind-Guess” AI Builders
Silent Routing Bugs in Field

Generic prompt-to-form generators convert text in 5 minutes by silently guessing whenever a Word doc has an incomplete programmer note (e.g., “Ask Q12 to core users”). Discovering broken skip logic or unclosed quotas after paying for 1,000 respondents ruins the dataset.

Option C // Resvanta AISP + Clarification Workspace
< 15 Minutes (QA-Verified)

AISP parses the `.docx` AST in ~30 seconds, pauses in an interactive Clarification Workspace so the researcher resolves every ambiguity in 2–5 minutes, and runs synthetic respondent bots to verify 100% of routing paths before launch.

02 // DETERMINISTIC 4-STAGE COMPILATION WORKFLOW

How Resvanta Compiles a Word (`.docx`) Survey in Under 15 Minutes

STAGE 01 // ~30 SECONDS

Direct OpenXML (`.docx`) AST Ingestion

Upload your raw `.docx` questionnaire directly. AISP parses document headings, tables, and bracketed programmer notes (`[SINGLE SELECT]`, `[TERMINATE IF Q2=4]`, `[PIPE Q3 SELECTED BRANDS]`, `[RANDOMIZE ROWS ANCHOR LAST]`) into a typed survey Abstract Syntax Tree (AST).

STAGE 02 // 2–5 MINUTES (HUMAN-IN-THE-LOOP)

Interactive Clarification Workspace (Zero Blind Guessing)

Every real-world Word questionnaire contains 2 to 6 underspecified rules—such as an unlinked quota cell, a missing scale anchor, or an ambiguous skip condition. Instead of guessing, AISP surfaces each item in the Clarification Workspace with 1-click deterministic options for researcher sign-off.

STAGE 03 // < 60 SECONDS

Quota Matrix, Piping & Econometric Module Compilation

Once clarifications are approved, the engine compiles nested cross-quotas (e.g., Age × Gender × Region), dynamic answer-list masking, Multinomial Logit (MNL) Conjoint / MaxDiff design matrices, and AirLock™ anti-fraud gates into an immutable survey specification.

STAGE 04 // ~3 MINUTES

Automated Synthetic Respondent QA & Live Link Deployment

Before a single human panelist enters the survey, AISP dispatches synthetic test respondents across every boolean branch to verify zero unreachable questions, zero infinite loops, and exact quota-full termination behavior—emitting a signed QA report and live fieldwork link.

03 // IMPLEMENTED CAPABILITIES, VALIDATION SAMPLE SIZE & SCOPE BOUNDARIES

Implemented Platform Capabilities, Benchmark Sample Size & Roadmap Scope

Resvanta Research Engine is currently available for Commercial Evaluation & Guided Pilot Deployments (`survey.resvanta.com`). Below is the exact distinction between capabilities implemented in the current release, the test protocol and sample size behind our numerical benchmarks, and roadmap items.

Implemented in Current Release vs. Roadmap
  • • Implemented Document Formats: Native OpenXML (`.docx`) and structured Markdown AST ingestion are live today. Cloud word-processor questionnaires are supported via standard `.docx` export. Direct multi-column PDF OCR ingestion is currently on the engineering roadmap.
  • • Implemented Question & Econometric Modules: Single/multi-select with exclusive options, numeric constraints, Likert/matrix grids, rank-order, Choice-Based Conjoint (MNL), MaxDiff, Van Westendorp, and Gabor-Granger pricing modules.
  • • Implemented Routing & Clarification Workspace: Multi-variable boolean skip/display logic (`AND`/`OR`/`NOT`), constructed list piping, interactive Clarification Workspace for ambiguity resolution, and synthetic respondent DAG reachability checks.
Benchmark Sample Size, Methodology & Scope Boundaries
  • • “< 15 Minute” Compilation Benchmark (Oct 2026 Internal Validation, n = 25): Evaluated across n = 25 standard B2B and consumer OpenXML (`.docx`) questionnaires (25–40 questions, containing 2–5 seeded ambiguity flags resolved in the Clarification Workspace). Median total time: 11.4 minutes (interquartile range: 8.5–14.2 minutes; min 6.8m, max 14.9m).
  • • “99.4% Bot & Script Rejection Rate” (Controlled Validation, n = 5,000 sessions): Measures the pre-survey gate rejection rate across 5,000 simulated headless-browser scripts, datacenter proxy requests, and duplicate device-hash submissions—not a claim that 99.4% of raw human panel traffic requires zero post-fieldwork cleaning.
  • • When Programming Takes 30–60+ Minutes: Multi-country tracking studies (100–150+ questions with `.xliff` translation overlays) or studies with custom JavaScript visual stimulus embeds require 30–60 minutes of researcher configuration after `.docx` ingestion.
04 // PUBLISHED LICENSING & EVALUATION PATHS

Transparent Annual Licensing for Research Teams & Agencies

Tier 0 Pilot ($2.5k) & Tier 1 Growth
US$18,000 / yr

Start with a US$2,500 Single-Study Pilot (1,500 completes, 100% credited within 30 days) or Tier 1 Corporate Growth at US$18,000/yr (25,000 Verified Completes/yr, 125k AirLock™ Screenings, 250 AISP Credits, and 250 MeritLock™ Starter Exams/yr).

* Software hosting capacity (panel sample billed at 0% markup)
Tier 2 // Corporate Enterprise (Track A)
US$54,000 / yr

Includes 150,000 Verified Completes/yr (750k AirLock™ Screenings), 1,500 AISP Credits (~60 studies), up to 5 Brand Workspaces, 15 Creator Seats, FRE 902(14) Court Certificates, Section 35 Clawback Dossiers, and 1,500 MeritLock™ Exams/yr.

* Track 3 Full RRE + MeritLock™ Enterprise Bundle: US$78,000/yr
Tier 3 // Agency Partner (Track B)
US$42,000 / yr Base

Multi-client bureau license with 15 isolated Client Sub-Workspaces, 50,000 Starter Verified Completes/yr, 1,000 AISP Credits, White-Label CNAME, 500 MeritLock™ Starter Exams, and 35% Wholesale Partner Discount on Client Complete Blocks.

* Track 3 Global Agency Full Ecosystem Bundle: US$95,000/yr
Tier 4 // Sovereign Dedicated VPC
US$165k / US$185k

Option 1 Client Self-Hosted BYOC (US$165,000/yr) or Option 2 Resvanta-Managed Dedicated Single-Tenant VPC (US$185,000/yr). Includes 1,000,000 Verified Completes/yr, 5,000 AISP Credits, 10,000 MeritLock™ Exams, and FIPS 204 WORM Ledger.

* MeritLock™ Standalone: $6.5k (Academic) / $18.5k (L&D) / $48k (Boards)
Test AISP with Your Own OpenXML (`.docx`) Questionnaire

Access the live platform at https://survey.resvanta.com or email [email protected] with a sample `.docx` questionnaire for a live 15-minute compilation walkthrough.

Open survey.resvanta.com