Automation Opportunity Audit

Next.jsTypeScriptOpenRouterKimiDeepSeekGrok 4.1 Fast

Start Here

Step 1
Describe Workflow Details
Tell us what your team does manually, step by step.
Step 2
Generate Audit Results
AI analyzes your workflow and estimates automation potential across key steps.
Step 3
Save to Portfolio
Portfolio View lets you rank and compare saved workflows.

Describe one workflow to audit

AI estimates current cost, savings range (P10/P50/P90), expected ROI timeline, and prioritized opportunities.

For best results, include:

Trigger
Process (steps, monthly volume, labor cost if available)
Systems
Rules
Result

Example Brief

Trigger: Customer support email arrives.
Process: Triage -> route -> draft -> QA -> send (~1,200 tickets/month at ~$45/hour blended).
Systems: Gmail, Zendesk, Notion KB.
Rules: Keep human approval on sensitive cases; no auto-send without review.
Result: Automate triage + first-response drafting while keeping CSAT stable.

OVERVIEW

A conversion-focused demo for AI automation services. Visitors describe a workflow in plain text and get an instant audit with P10/P50/P90 savings, expected ROI timeline, prioritized opportunities, confidence/method context, and a portfolio view to rank and compare saved audits.

ARCHITECTURE

The app combines a client-side audit UI with a Next.js API route. The backend builds structured prompts, calls OpenRouter with a primary model plus fallback routing, validates and normalizes JSON, then applies deterministic scoring and derivation layers before rendering. It includes guardrails, input-quality warnings, and section-by-section streaming for progressive loading. A client-side portfolio layer stores saved audit snapshots in localStorage and computes ranked portfolio views and side-by-side workflow comparisons.

FUNCTIONALITY

Free-text workflow intake with no required form complexity
Single-field intake flow to reduce friction and increase completions
Model routing via OpenRouter with primary + fallback support
Schema validation and normalization before UI rendering
P10/P50/P90 monthly savings outputs for uncertainty-aware estimates
Expected ROI timeline derived from implementation cost and savings bands
Task-level automation breakdown and prioritized top opportunities
Confidence badges, derivation logic tables, and edge-case warnings
Portfolio View with save, rank, compare, and delete workflow snapshots
Client-side session persistence for portfolio snapshots via localStorage
Server-side rate limiting (5/min and 30/day) per IP and session/user

HOW IT WORKS

A user submits workflow details from the project page. The API builds a structured audit prompt, calls the selected OpenRouter model, and enforces strict JSON output. The response is parsed, validated, and run through deterministic scoring/derivation logic before being streamed into grouped UI sections. Users can then save any result to Portfolio View, where snapshots are ranked by weighted ROI and execution signals and compared side by side. If provider output fails, the route applies fallback handling and returns explicit warning context.

OUTCOMES

Creates a low-friction bridge from anonymous visitor to qualified lead conversation.
Demonstrates practical automation reasoning with structured, explainable outputs.
Gives prospects immediate value while making assumptions explicit for follow-up scoping.
Enables multi-workflow prioritization with a session-based portfolio ranking and compare workflow.
Protects API spend and abuse surface with layered server-side rate limits.