Redesign products and workflows for the AI era.

Most AI redesigns skip the decisions that matter: which steps should stay rules, which ones a model should own, and who approves an action before it happens. ProductizeAI-Architect makes those decisions explicit, so a team can argue with them before building anything.

Paste a product description, PRD or workflow (or attach a readable file) and say "Productize this." The redesign is built around four decisions.

What you get

By default, a quick pass: one workflow and up to three changes. Ask for the full blueprint and you get the before/after experience, a mechanism and autonomy map, failure recovery, an evaluation plan, a cost model and the first experiment to run. Every redesign ends with the unresolved question that could reverse the recommendation.

The output separates what you told it from what it inferred and what it assumed. It doesn't invent ROI, adoption numbers or validated thresholds.

Who it's for

Product managers, founders and product leaders deciding where AI belongs in an existing product or process.

Get it

Install in Claude: go to Customize > Skills, click +, choose Create skill, then Upload a skill and select the zip. "Code execution and file creation" must be on in Settings > Capabilities. In Claude Code: unzip it into ~/.claude/skills/.

Try it

Productize our onboarding workflow.

Redesign this SaaS workflow with AI and show what should stay deterministic.

Should this feature be a copilot, an agent or ordinary automation?

The more concrete the input (who does each step, how often, what goes wrong today), the more specific the redesign. If the product or workflow is missing, ProductizeAI-Architect asks for it before answering.

Limits

• It works only on what you share in the conversation. It has no connectors, doesn't access your tools or analytics, and takes no actions in any system.
• It can analyze a website only if the assistant you run it in (ChatGPT or Claude) can retrieve the page. If it can't, it will ask you to paste the content.
• For high-impact decisions it keeps consequential judgment with an accountable human, and flags where expert review is needed.
• The output is a design proposal to review with your team. A complete document is not the same as production readiness.

ProductizeAI-Architect is free to use.

Questions, bug reports and feedback go to one inbox, which I read myself.

I aim to reply within 3 business days. To help me understand the problem, include:

• What you asked ProductizeAI-Architect to do, and what came back
• What you expected instead
• Roughly when it happened

Please don't send confidential material. A short, anonymized example is enough.

Common questions

The redesign is too generic. Add specifics: the target user, who does each step today, volumes, costs and what has gone wrong. ProductizeAI-Architect reasons from what you give it, and it labels its assumptions so you can see where it had to guess.

It says it couldn't read my link. Website analysis depends on ChatGPT or Claude being able to retrieve the page. Paste the relevant content instead.

Can it connect to Jira, my analytics or my CRM? Not in this version. It's an analysis tool with no integrations.

How do I remove it? Disable or remove the plugin in ChatGPT, or the skill in Claude, at any time.

This policy covers ProductizeAI-Architect ("the plugin"), created by Irene Pylypenko in Montreal, Quebec, Canada ("I", "me").

In short: I don't collect any data about your use of the plugin. ChatGPT and Claude process your conversations under OpenAI's and Anthropic's policies. The only personal data I ever receive is what you send me if you email for support.

1. I don't collect plugin usage data

ProductizeAI-Architect is a set of instructions that runs entirely inside ChatGPT or Claude. It has no server, database, connectors or accounts, so there is nowhere for your data to go. Specifically, the plugin does not:

• receive, read or store your prompts, uploaded files or the responses ChatGPT or Claude generates
• use analytics, telemetry, cookies, tracking pixels or device identifiers
• log who uses it, when, or how often
• sell, share or use any data for advertising or model training

If OpenAI makes aggregate statistics available to plugin publishers, such as install counts, I may view them. These don't identify you or reveal what you wrote.

2. ChatGPT processes your interactions under OpenAI's policies

When you use ProductizeAI-Architect, your conversation, including text and files you share and any web pages ChatGPT retrieves for you, is processed by OpenAI as the operator of ChatGPT. How OpenAI collects, uses, retains and protects that data is governed by OpenAI's privacy policy and the data controls in your ChatGPT settings, not by me. I have no access to your ChatGPT conversations.

To manage, export or delete your conversations, or to control whether they're used to improve OpenAI's models, use the data controls in ChatGPT.

If you use the Claude skill, the same applies with Anthropic: your conversation is processed by Anthropic under Anthropic's privacy policy and your Claude settings, and I have no access to it. Downloading the skill from this site or GitHub doesn't send me any information about you.

ProductizeAI-Architect doesn't need personal information to work. Sanitized or anonymized material produces the same quality of redesign, and I recommend not sharing other people's personal data, health information, payment card data, government ID numbers or passwords.

3. How I handle support emails

If you email the support address, I receive your name, email address, your message and anything you attach.

• Use: only to reply to you and fix the problem you report. I don't add you to a mailing list or use your email for marketing.
• Storage: in my email account, hosted by Google.
• Sharing: with no one, unless required by law.
• Retention: as long as needed to resolve your request, then deleted within 12 months.
• Your rights: you can ask me to access, correct or delete your correspondence at any time. I'll respond within 30 days.

Please don't include confidential material in support emails. A short, anonymized example is enough.

Children

ProductizeAI-Architect is a professional tool and isn't directed at children.

Person responsible for personal information

Irene Pylypenko is responsible for the protection of personal information under Quebec's Act respecting the protection of personal information in the private sector. Contact: the support address.

Changes

If a future version of the plugin collects data (for example, through a server or integrations), this policy will be updated before that version is released, and the effective date above will change.

These terms apply when you use ProductizeAI-Architect. By using it, you agree to them. Your use of ChatGPT is also subject to OpenAI's terms and policies, and your use of Claude to Anthropic's terms and policies.

The service

ProductizeAI-Architect provides a methodology that guides ChatGPT or Claude to produce AI product redesign proposals from material you supply. I may change, suspend or discontinue it at any time.

Your content and output

You keep all rights to what you share and to the output you receive. You're responsible for having the right to share your input, including anything confidential to your employer or clients.

The methodology

The ProductizeAI-Architect methodology, instructions and reference material are © 2026 Irene Pylypenko.

Acceptable use

Don't use ProductizeAI-Architect to break the law, infringe anyone's rights, process other people's personal data without authorization, or violate OpenAI's or Anthropic's usage policies.

No professional advice

Output is AI-generated and can be incomplete or wrong. It's a starting point for your own judgment, not legal, regulatory, security, financial or other professional advice. Cost estimates and success criteria are proposals, not validated results. You're responsible for reviewing the output and for decisions you make with it.

Disclaimer

ProductizeAI-Architect is provided "as is" and "as available", without warranties of any kind, express or implied, including accuracy, fitness for a particular purpose and non-infringement, to the extent permitted by law.

Limitation of liability

To the extent permitted by law, I am not liable for any indirect, incidental, special, consequential or punitive damages, or for loss of profits, revenue, data or business, arising from your use of ProductizeAI-Architect. Nothing in these terms limits liability that cannot be limited under applicable law, including the law of Quebec.

Governing law

These terms are governed by the laws of the Province of Quebec and the federal laws of Canada that apply there. Disputes will be heard by the courts of the judicial district of Montreal, unless the law where you live gives you the right to bring them elsewhere.

Changes

I may update these terms. The effective date above shows when they last changed. Continuing to use ProductizeAI-Architect after a change means you accept the updated terms.

Contact

Use the support address.