Methodology & transparency

We are not magic.
Here is exactly what we do.

Pulse is an AI-assisted audit tool. It analyses your project code systematically, applies a documented scoring model, and produces structured documents. This page explains every step — what the score means, how the budget is calculated, and where our estimates can be wrong.

What you get

Free
€0
Always free — no account needed
Verdict — one plain-English sentence on your project state
Issue matrix — security × infrastructure, by severity
3 headline numbers — critical, fixable in Lovable, need a developer
Green impact % — how much risk you can eliminate yourself
Pro
€19/month
Unlimited audits · Cancel anytime
Everything in Free — plus the full action plan
Full fix list — every issue with 🟢🟠🔴 feasibility signal
Lovable prompts — ready-to-paste for every green issue
Developer estimates — days + cost for every red issue

How Pulse analyses your project

Your project is analysed by two independent AI agents running in parallel. Each agent is a specialist — it only looks at its own domain and finds concrete issues with file references and severity levels. A third agent then takes all issues and produces a prioritised action plan.

Security agent
🔒 Security
Authentication & authorisation gaps
Secrets and API key exposure
Row-level security policy completeness
Input validation on API routes
CORS configuration & data exposure
Infrastructure agent
🏗️ Infrastructure
Hosting & deployment setup
Database backup & resilience
Scalability risks & bottlenecks
Monitoring, logging, observability
Environment variable management

Code quality and handover readiness are intentionally out of scope. We focus on what can break, expose user data, or cause the app to go down — because those are the issues that matter before you scale or hand off.

Every issue is assessed for how easy it is to fix — specifically, whether it can be fixed directly in Lovable with a prompt, or whether it requires a developer.

🟢
Fix in Lovable
Can be resolved with a specific prompt to Lovable. We provide a ready-to-paste prompt for each green issue. Examples: enabling RLS on a table, adding an auth check to a route, setting environment variables correctly.
🟠
Needs review
Partially fixable in Lovable but requires careful testing or developer oversight. Fixing blindly with a prompt risks breaking other things. Examples: CORS changes, rate limiting, auth flow adjustments.
🔴
Requires a developer
Structural issue — cannot be safely fixed with an AI prompt. Includes hosting migration, database schema redesign, CI/CD setup, and infrastructure changes. We provide a day estimate and cost range.

How the green impact % is calculated

The green impact percentage tells you how much of your critical risk you can eliminate yourself, today, without a developer. It is calculated deterministically — no AI involved.

greenImpact = round( greenPoints / totalPoints × 100 )
Where each issue is weighted by severity: Critical = 3 pts · High = 2 pts · Medium = 1 pt
Green points = sum of severity weights for all 🟢 issues. Total points = sum of severity weights for all issues. A project with 2 critical green issues and 1 critical red issue would score 60% (6 green pts out of 10 total pts).

This score is purely mathematical — it reflects the relative severity of what you can fix yourself versus what requires professional help. A high green impact means most of your risk is self-serviceable. A low score means the structural issues are the heaviest ones.


How developer cost estimates work

For every 🔴 issue that requires a developer, Pulse estimates the number of days and the developer profile needed. The cost is then calculated deterministically — no AI guesswork.

Backend
€800/day
API, database, business logic
Frontend
€700/day
UI, components, integration
DevOps
€600/day
Infra, CI/CD, deployment
Cost = devDays × rate (per profile)
The AI agent estimates the number of days. The cost calculation is deterministic JavaScript — no AI involved.
Rates are 2025 market averages for senior freelancers or mid-market agencies in Western Europe. Always confirm against real quotes.

What Pulse assumes — and where it can be wrong

The day rates above are 2024–2025 market averages for senior freelancers or mid-market agencies in Western Europe. Your actual cost will depend on who you hire and where they are based.

ProfileAdjustmentTypical context
Western Europe (France, Germany, Netherlands) Reference Our baseline — senior freelancer or mid-market studio
Eastern Europe (Poland, Romania, Bulgaria) −20% to −30% Strong talent pool, increasingly common for remote-first projects
Junior or offshore teams −30% to −50% Lower cost, but add time for onboarding and quality review
Specialised agencies (security, fintech, healthcare) +30% to +50% Domain expertise commands a premium
Large consultancies (Big 4, Capgemini, etc.) +60% to +100% Significant overhead; rarely the right fit for early-stage refactoring
What a Pulse audit does not replace
A real quote from a development studio
A human senior engineer code review
A certified security penetration test
A legal or compliance assessment
A performance or load test
Architecture consulting for complex systems
Our honest disclaimer
🤖
Pulse is AI-assisted, not AI-perfect. Our agents analyse the code you provide and apply a documented scoring model. They can miss context, misinterpret patterns, or overlook issues that a human reviewer would catch. We catch common errors with a QA pass, but we do not guarantee completeness.
📁
The analysis is limited to what you upload. If your ZIP export is incomplete, if key files are excluded, or if the project structure is atypical, the audit quality will be affected. The tool tells you which files it read — if something important is missing, the score may not reflect reality.
💶
The budget is an estimate, not a quote. Our roadmap and budget figures give you a realistic order of magnitude to start conversations with developers. They are not a binding commitment. Always request formal quotes from at least two studios before making a financial decision.
🕐
Timelines depend on team velocity. A Phase 1 estimated at 2 weeks assumes a focused, experienced team working on this full-time. Part-time involvement, onboarding time, or scope changes will affect the actual duration.
Ready to audit your project?
Free executive summary — no account required
Run an audit →