WalnutAI — From Requirements to Code to Tests, Automatically

Upload any document — Word, PDF, Excel, CSV, or PowerPoint — and WalnutAI automatically extracts a structured hierarchy of Epics, Features, Modules, and User Stories with acceptance criteria. What used to take days of BA effort takes minutes.

Already have code but no requirements? WalnutAI analyzes your entire codebase — folder structure, AST parsing, module boundaries — and generates a complete requirements hierarchy from the actual implementation. Perfect for legacy systems and undocumented codebases.

Import user stories and work items from Jira or Azure DevOps with custom field mapping and bidirectional sync. WalnutAI keeps your stories aligned with your project management tools automatically — no double-entry, no drift.

Connect directly to Figma and select screens, frames, or flows. WalnutAI’s vision AI reads your designs and generates user stories based on the UI behaviors and interactions it sees — bridging the gap between design and development.

Assign any user story to a WalnutAI Cloud Agent. The agent reads your codebase, understands the requirements, implements the feature, commits to a feature branch, and raises a draft pull request — autonomously. You review. The agent ships.

Log a defect in WalnutAI’s workboard and the AI agent reads the failure context — test case, steps, screenshots, expected vs actual — finds the broken code, and raises a fix PR automatically. No copy-pasting into your editor. No re-explaining what QA already documented.

Bring WalnutAI into your editor. The VS Code extension puts Ask, Plan, and Agent modes directly in your IDE — query your codebase, draft implementation plans with human approval, and spawn autonomous sub-agents without leaving VS Code.

Connect external MCP servers — Figma, GitHub, databases, internal APIs, and more — directly into WalnutAI’s code editor. Your AI agents can call any connected tool as they work, so nothing needs to be copy-pasted in from another tab.

WalnutAI’s three-phase gap analysis engine compares your user stories against your codebase to find exactly what’s missing, incomplete, or outdated. Phase 1 finds code gaps. Phase 2 validates the stories themselves. Phase 3 scores code quality per story.

Generate complete, step-by-step test cases — positive, negative, API, UI, and integration scenarios — from any user story in under 30 seconds. WalnutAI produces a human-reviewable draft for every story. Teams cut test creation time by up to 40%.

Interact with your web or mobile app and WalnutAI converts every click, input, and navigation into a structured test case automatically. No scripting. No transcription. Works for web, Android, and iOS — test cases ready the moment you stop recording.

An interactive, always-current visual map linking every Epic → Feature → User Story → Test Case → Defect. Click any node for full details. Export to Excel for compliance audits. Know your test coverage at a glance — automatically maintained every sprint.

Execute tests step-by-step, capture screenshots for failures, track pass/fail/blocked/skip results, and generate rich PDF reports — all in one place. Suite-level reports give leadership real-time quality visibility at release time.

Track test execution trends, defect rates, flaky test detection, team productivity, and AI model token usage — across every project and sprint. WalnutAI’s analytics dashboards turn QA data into release confidence and executive-ready ROI reporting.

Choose any AI model — OpenAI, Azure OpenAI, AWS Bedrock (Claude), Google Gemini, Anthropic, or local Ollama. Per-project model configuration with AES-256 encrypted keys and spending limits gives enterprises full control over AI cost, privacy, and performance.

Enterprise-grade RBAC with granular permissions across organizations, projects, and roles (Admin, Manager, Tester, Developer, Viewer). Immutable audit logs with 7-year retention keep every story and test action permanently recorded for compliance.