Skip to main content
Documentation

Frantiko documentation

Everything you need to plan, build, and ship better products with AI. Use the navigation above or jump straight to a guide below.

Quick start

Create your workspace and generate your first blueprint in under 5 minutes.

Core concepts

Understand blueprints, sessions, fix requests, and health reports.

AI Architect

Generate structured product blueprints from a short product description.

Problem Discovery

Scan markets for real, ranked opportunity gaps before you start building.

Market Validation

Get an AI success score and a Build or Pivot verdict for any idea.

Session Insights

Understand how real users navigate your product using first-party session data.

AI Fix Center

Turn session friction and bugs into ready-to-paste prompts for your AI builder.

Revenue Advisor

Get pricing strategy, conversion playbook, and investor pitch generation.

Integrations

Connect Frantiko to Bolt, Cursor, Lovable, v0, GitHub, Slack, and more.

Quick start

#

This guide takes you from a new account to your first blueprint and live session data in under 5 minutes.

1. Create your account

Go to frantiko.me/auth/signup and enter your email address and a password to create your account. No credit card is required.

2. Create a project

From your dashboard, select New project. Give it a name — for example, My SaaS App — and a short description of what the product does.

Your project is the container for all blueprints, sessions, fix requests, and reports.

3. Install the Frantiko tracking snippet

To track user sessions, add the Frantiko snippet to your product. Go to Dashboard → Integrations → Frantiko Snippet and copy your unique project token.

Paste the following snippet inside the <head> of your HTML, replacing YOUR_PROJECT_TOKEN with the token you copied:

<!-- Frantiko Session Insights -->
<script>
  (function(f,r,a,n,t){
    f[t]=f[t]||function(){(f[t].q=f[t].q||[]).push(arguments)};
    var s=r.createElement('script');s.async=1;s.src=a;
    var x=r.getElementsByTagName('script')[0];
    x.parentNode.insertBefore(s,x);
  })(window,document,'https://cdn.frantiko.me/tracker.js','','frantiko');
  frantiko('init', 'YOUR_PROJECT_TOKEN');
</script>

4. Generate your first blueprint

Go to AI Architect in the sidebar. Describe your product in two to four sentences. Click Generate blueprint. Within 15 to 30 seconds, Frantiko produces a structured blueprint with features, user stories, and a recommended tech stack.

You are ready to build

Your blueprint, session data collection, and project workspace are now active. Start building and come back to Session Insights once your first users land.

Core concepts

#

Blueprints

A blueprint is a structured product specification generated by AI Architect. It contains a product summary, core feature list, user stories, data model outline, and recommended stack. You can export any blueprint as a PDF to share with collaborators or paste into your AI builder.

Sessions

A session is one complete visit by a user to your product, captured by the Frantiko tracking snippet. Sessions include page views, click events, scroll depth, rage clicks, and dead clicks. Frantiko aggregates these into the Session Insights view so you do not need to watch every recording individually.

Fix requests

A fix request is a structured prompt generated by AI Fix Center from a reported issue or session signal. Each fix request includes the issue description, reproduction steps, affected components, and a ready-to-paste prompt for your AI builder (Bolt, Cursor, Lovable, or v0).

Health reports

A health report is a weekly summary of your product's key metrics: active sessions, fix request volume, blueprint progress, top user flows, and a health score out of 100. Reports are generated automatically every Monday and can be exported as a PDF.

AI Architect

#

AI Architect generates a complete product blueprint from a short description. You use the blueprint as the foundation for prompting your AI builder and keeping your product scope defined as you iterate.

How to generate a blueprint

  1. Open AI Architect from the dashboard sidebar.
  2. In the input field, describe your product. Include the target user, the core problem it solves, and one to three key features.
  3. Select the Builder you are using (Bolt, Cursor, Lovable, v0, or Other).
  4. Click Generate blueprint. The generation takes 15 to 30 seconds.
  5. Review the blueprint. Use the Refine button to add more context and regenerate any section.
  6. Click Export PDF to download or Copy prompt to paste directly into your builder.

Blueprint structure

Every generated blueprint contains the following sections:

Product summaryOne-paragraph description of the product, its users, and its core value.
Core featuresNumbered list of the main features with a one-line description of each.
User storiesAs a [user], I want to [action] so that [outcome] — for each feature.
Data modelKey entities and their relationships in plain language.
Tech stackRecommended stack based on your selected builder and feature requirements.
Build orderSuggested order to build features to minimize rework and dependencies.

Note

Blueprints are AI-generated and should be reviewed before use. The AI may suggest technologies or patterns that are not appropriate for your specific context. Use the Refine feature to correct any section.

Problem Discovery

#

Problem Discovery scans a market or product space for real, ranked opportunity gaps before you start building. It surfaces unmet needs, underserved user segments, and recurring complaints from public data — so you can validate direction without manual research.

How to run a discovery

  1. Open Problem Discovery from the dashboard sidebar.
  2. Enter a market, product category, or audience. For example: no-code app builders, solo founders, or remote team collaboration tools.
  3. Optionally, specify a geographic focus or a specific pain theme.
  4. Click Discover problems. The AI scans public signals and synthesizes a ranked list of problems.
  5. Each result includes a problem summary, estimated frequency, urgency score, and a suggested product angle.

Discovery result structure

Problem titleConcise name for the unmet need or friction point.
DescriptionTwo to three sentences explaining the problem and who experiences it.
FrequencyHow often this problem appears across public signals: low, medium, or high.
UrgencyHow acutely users feel this problem: nice-to-have, important, or critical.
Product angleA suggested way to address the problem with a product or feature.
Related searchesKeywords and phrases associated with this problem for further research.

Tip

Run Problem Discovery before Market Validation. Use a discovery result as the starting idea for a Market Validation score — the workflow is designed to chain these two tools together.

Market Validation

#

Market Validation gives any product idea an AI success score and a clear Build or Pivot verdict. It evaluates competition, market size, uniqueness, timing, and execution complexity — and explains the reasoning behind every score.

How to validate an idea

  1. Open Market Validation from the dashboard sidebar.
  2. Describe your product idea in two to five sentences. Include the target user, the core problem, and how your solution is different from existing options.
  3. Click Validate idea. The AI evaluates your idea across five dimensions and produces a score from 0 to 100.
  4. Review the verdict (Build, Validate further, or Pivot) and the reasoning for each dimension.
  5. Use the Pivot suggestions section to explore adjacent angles if the verdict is Pivot.

Validation dimensions

Market sizeEstimated total addressable market and whether it is large enough to build a sustainable business.
CompetitionHow many established players exist, how entrenched they are, and whether there is room to compete.
DifferentiationHow clearly the idea differs from existing solutions — a score of 1 means commodity, 10 means truly unique.
TimingWhether the market is ready for this solution now, too early, or past its growth window.
Execution riskHow hard this is to build given the required technology, regulatory context, and team size.

Verdict guide

Build

Score 70–100

Strong market signal, clear differentiation, good timing. Move to AI Architect and start building.

Validate further

Score 40–69

The idea has merit but one or more dimensions need more evidence. Run a Problem Discovery search to sharpen the angle before building.

Pivot

Score 0–39

Significant market, competition, or differentiation issues. Review the Pivot suggestions section and try a revised angle.

AI Fix Center

#

AI Fix Center converts product issues — reported manually or surfaced from session data — into structured prompts you can paste into your AI builder to resolve them.

How to create a fix request

You can create a fix request in two ways: manually or automatically from a Session Insights friction signal.

Manually

  1. Open AI Fix Center from the dashboard sidebar.
  2. Click New fix request.
  3. Give the issue a short title, then describe it in the error description field. You can paste a user report, a console error, or write a plain-language description.
  4. Optionally add reproduction steps and select a severity level.
  5. Click Generate AI fix. Frantiko produces a structured diagnosis and a ready-to-paste prompt for your builder.
  6. Copy the fix prompt and paste it into Bolt, Cursor, Lovable, or any other AI builder.

Automatically from Session Insights

  1. When Frantiko detects a friction signal, it creates a dashboard notification with a summary of the affected element and user count.
  2. Click Open in AI Fix Center inside the notification. The fix request form opens with the title, description, and session context already filled in.
  3. Review the pre-filled content, adjust if needed, and click Generate AI fix.

Fix prompt structure

Each generated fix prompt follows this structure to maximize clarity for your AI builder:

## Issue
[Clear one-sentence description of what is broken]

## Expected behavior
[What the user expected to happen]

## Actual behavior
[What is actually happening]

## Steps to reproduce
1. [Step one]
2. [Step two]
3. [Step three]

## Fix instruction
[Specific instruction for the AI builder: component, file, or behavior to change]

Session Insights

#

Session Insights gives you a behavioral view of your users pulled from the Frantiko tracking snippet. You do not need to watch individual session recordings — Frantiko surfaces the most important signals automatically.

Available signals

Rage clicksElements that users click repeatedly in frustration — a strong indicator of broken or confusing UI.
Dead clicksClicks on elements that do not respond — often non-interactive items that users expect to be clickable.
Drop-off pointsPages or steps where users leave the product mid-flow.
Scroll depthHow far users scroll on key pages — useful for content and landing page optimization.
Top user pathsThe most common navigation sequences through your product.
Session heatmapAggregated click and attention map for any page.

Auto-detect, auto-notify, auto-direct

Session Insights runs a continuous background pipeline that monitors your session data and acts without manual intervention.

Auto-detectFrantiko scans incoming session events every hour for friction signals: rage clicks above a threshold, drop-off spikes, and dead click clusters. When a pattern exceeds the threshold, it is flagged automatically.
Auto-notifyWhen a friction signal is detected, Frantiko creates a dashboard notification with a plain-language summary: which element, how many users affected, and the estimated impact. If Slack or Discord is connected, the alert is also sent there.
Auto-direct to fixEach friction notification has an Open in AI Fix Center button. Clicking it pre-fills the fix request form with the element description, affected user count, and the session context — so you can generate a fix prompt in one click without re-describing the issue.

Installing the Frantiko tracking snippet

To collect session data, add the Frantiko snippet to your product. You can find your unique tracking key in Dashboard → Session Insights → Setup.

<script>
  (function(f,r,a,n,t,i,k,o){
    f[n]=f[n]||function(){(f[n].q=f[n].q||[]).push(arguments)};
    i=r.createElement(a);i.async=1;i.src=t;
    k=r.getElementsByTagName(a)[0];k.parentNode.insertBefore(i,k);
  })(window,document,'script','frantiko','https://cdn.frantiko.com/tracker.js');
  frantiko('init', 'YOUR_TRACKING_KEY');
  frantiko('trackPageview');
</script>

Place the snippet before the closing </body> tag. Replace YOUR_TRACKING_KEY with the key shown in your Session Insights setup panel.

Requirements

  • The Frantiko tracking snippet installed on your product (see Integrations).
  • At least one active session recorded — usually within 24 hours of installation.
  • The snippet must be installed for auto-detect and auto-notify to function.

Product Reports

#

Product Reports give you a weekly summary of your product health. Use them to track progress over time, communicate status to stakeholders, or do a retrospective at the end of a sprint.

Report contents

Health scoreA score from 0 to 100 calculated from session quality, fix rate, and blueprint completion.
Weekly sessionsTotal sessions recorded in the last 7 days with a trend indicator.
Fix requestsNumber of fix requests opened, resolved, and pending this week.
Blueprint progressHow many blueprint features have been shipped versus remaining.
Top issuesThe three most-reported or session-triggered issues this week.
RecommendationsAI-generated suggestions based on the week's data.

Exporting a report

  1. Open Product Reports from the dashboard sidebar.
  2. Select the report you want to export from the list.
  3. Click Export PDF. The PDF includes your report title, all metric sections, and a Frantiko-branded footer with the generation date and page numbers.

Revenue Advisor

#

Revenue Advisor generates a complete monetization plan for your product — pricing model, conversion playbook, and a short investor pitch — based on a description of what you are building and who you are building it for.

How to get a revenue plan

  1. Open Revenue Advisor from the dashboard sidebar.
  2. Describe your product: what it does, who it is for, how it is different, and your current stage (idea, MVP, or live product).
  3. Click Analyze revenue strategy. The AI produces a structured plan tailored to your product and market context.
  4. Review each section: recommended pricing model, suggested tiers and prices, conversion tactics, and the investor pitch paragraph.
  5. Use the copy button on the investor pitch to export it for a deck or cold email.

Revenue plan contents

Pricing modelThe recommended monetization approach: freemium, seat-based, usage-based, flat subscription, or one-time. Includes the reasoning behind the recommendation.
Pricing tiersTwo to three concrete tiers with suggested names, price points, and what each tier includes. Grounded in your described market and audience.
Free plan scopeIf freemium is recommended: what to include in the free tier to drive activation without undermining conversion.
Conversion tacticsSpecific techniques to move users from free to paid — upgrade prompts, limit triggers, and feature gates — tailored to your product type.
Annual discountWhether to offer annual billing, at what discount, and how to frame it on your pricing page.
Investor pitchA two to three sentence pitch paragraph describing your product, market, and traction hook. Ready to paste into a deck or email.

When to use this

Run Revenue Advisor after you have a validated idea (Market Validation score above 50) and before you write your landing page copy. The pricing tiers and positioning language it generates can be used directly in your pricing section.

Integrations

#

Frantiko connects to the tools in your workflow so data flows in and actions flow out. All integrations are configured from Dashboard → Integrations. Most require only an API key or webhook URL — no code changes on your product.

Setting up the Frantiko snippet (required for Session Insights)

The Frantiko tracking snippet is the first-party behavioral data layer that powers Session Insights. Install it on your product and enter your project token in Frantiko before session data will appear.

  1. In Frantiko, go to Dashboard → Integrations → Frantiko Snippet and copy your project token.
  2. Add the following script tag to the <head> of your product, replacing YOUR_PROJECT_TOKEN:
<script>
  (function(f,r,a,n,t){
    f[t]=f[t]||function(){(f[t].q=f[t].q||[]).push(arguments)};
    var s=r.createElement('script');s.async=1;s.src=a;
    var x=r.getElementsByTagName('script')[0];
    x.parentNode.insertBefore(s,x);
  })(window,document,'https://cdn.frantiko.me/tracker.js','','frantiko');
  frantiko('init', 'YOUR_PROJECT_TOKEN');
</script>
  1. Click Save configuration in the Frantiko Integrations panel. Session Insights will start showing data within 24 hours of your first recorded session.

All integrations

Frantiko Snippet
Required
Session data source. Powers Session Insights, friction detection, and the auto-notify pipeline.
Bolt.new
Optional
Errors from your Bolt project are forwarded to AI Fix Center automatically. No manual copy-paste.
Lovable
Optional
Connects your Lovable project so UX signals and errors reach Frantiko in real time.
GitHub
Optional
Open a pre-filled GitHub issue from any AI Fix Center result. Also links blueprints to your repo.
Vercel
Optional
Correlates deployments with error spikes. See exactly which deploy introduced a regression.
Linear
Optional
Create a Linear issue from any friction signal or fix with one click. Priority and description pre-filled.
Slack
Optional
Sends a notification when Frantiko detects a new friction signal or completes a report.
Discord
Optional
Route Frantiko alerts to a Discord server for indie hacker-style communities.
Notion
Optional
Sync blueprints, market validation reports, and weekly digests to your Notion workspace.
Zapier
Optional
Trigger a Zap on any Frantiko event: friction detection, report completion, or fix generation.

Frequently asked questions

#

Do I need to know how to code to use Frantiko?

No. Frantiko is designed for builders who use AI code editors like Bolt, Cursor, and Lovable. You describe your product in plain language and Frantiko generates the structure, blueprints, and prompts. You never write code inside Frantiko.

What AI model does Frantiko use?

Frantiko uses Google Gemini 2.5 Flash as its primary model for blueprint generation, market validation, and fix prompt creation. If the primary model is under load, it falls back to Gemini 2.0 Flash automatically. The model selection is handled for you.

Where is my data stored?

Your blueprints, fix requests, reports, and account settings are stored in a PostgreSQL database hosted on Neon (AWS us-east-1). Session data is collected first-party via the Frantiko tracking snippet and stored in the same Neon database — no third-party behavioral analytics vendor is involved.

Is the Frantiko snippet required?

The Frantiko tracking snippet is required only if you want Session Insights, auto-detect friction signals, and auto-notify alerts. All other Frantiko features — AI Architect, Problem Discovery, Market Validation, AI Fix Center, and Revenue Advisor — work without it.

How does auto-detect work?

Frantiko processes incoming session events every hour for friction signals. When a pattern exceeds the threshold — such as rage clicks above five events on the same element, or a page drop-off rate above 60% — it creates an automatic notification and pre-fills an AI Fix Center request with the relevant context. You can then generate a fix prompt in one click.

Can I connect my own Bolt or Lovable project?

Yes. Connect your project token from Dashboard → Integrations. Once connected, error events from your builder are forwarded to Frantiko automatically and appear as pre-filled fix requests in AI Fix Center.

How does billing work?

Frantiko uses Gumroad for payments. On the Billing page you will find links to the Pro and Teams plan checkout pages on Gumroad. After completing your purchase, Gumroad sends you a license key — paste it into the Billing page to activate your plan instantly. No credit card details are stored by Frantiko.

How does the Free plan differ from Pro?

Free includes 3 blueprints per month, 1 project, 5 Problem Discovery searches, and 3 Market Validation scores. Pro ($19/month via Gumroad) unlocks unlimited blueprints, fix prompts, Problem Discovery, Market Validation, Revenue Advisor, up to 10 projects, and 90-day session history. Team ($59/month) adds unlimited projects, unlimited team members, role-based access, and collaborative workspaces.

Is there a limit on blueprint size?

Blueprint generation is limited by the model output window. For very large products we recommend splitting the product into modules and generating one blueprint per module. Each blueprint can hold up to 20 pages, 20 database tables, and 30 API routes.

Can I delete my data?

Yes. Delete individual blueprints, fix requests, and reports from the dashboard. To delete your entire account and all associated data, go to Settings → Account → Delete account. Deletion is immediate and cannot be undone.

Does Frantiko work for non-SaaS products?

Yes. Frantiko works for any digital product — mobile apps, marketplaces, content platforms, and internal tools. The AI Architect adapts the blueprint structure to the product type you describe. Problem Discovery and Market Validation work for any market.

Still have questions?

Reach the team on Discord or by email.

DiscordContact us