<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Written reviews on Nefer Lopez</title><link>https://neferlopez.com/reviews/</link><description>Recent content in Written reviews on Nefer Lopez</description><generator>Hugo</generator><language>en-US</language><atom:link href="https://neferlopez.com/reviews/index.xml" rel="self" type="application/rss+xml"/><item><title>Inline Variable Formatting Is Typebot's Most Underrated Feature</title><link>https://neferlopez.com/reviews/typebot-inline-variable-formatting/</link><pubDate>Fri, 24 Nov 2023 00:00:00 +0000</pubDate><guid>https://neferlopez.com/reviews/typebot-inline-variable-formatting/</guid><description>&lt;p&gt;Most people use Typebot like a flow chart: one block per idea, variables set here, displayed there, a &amp;ldquo;Script&amp;rdquo; block whenever the logic gets spicy. That works — until your flow is forty blocks of scaffolding around three lines of actual logic.&lt;/p&gt;
&lt;p&gt;The &lt;a href="https://www.youtube.com/watch?v=8Il-SEJplYY"&gt;video&lt;/a&gt; covers the feature that changed how I build: &lt;strong&gt;inline variable formatting&lt;/strong&gt; — JavaScript, evaluated directly inside any text block.&lt;/p&gt;
&lt;h2 id="what-it-actually-is"&gt;What it actually is&lt;/h2&gt;
&lt;p&gt;Anywhere you can type a variable in Typebot, you can also type an expression. That means you can:&lt;/p&gt;</description></item><item><title>Stop Paying the Middleman: OpenAI Vision Straight From Typebot</title><link>https://neferlopez.com/reviews/openai-vision-direct-typebot/</link><pubDate>Sun, 19 Nov 2023 00:00:00 +0000</pubDate><guid>https://neferlopez.com/reviews/openai-vision-direct-typebot/</guid><description>&lt;p&gt;Every integration tutorial starts the same way: &amp;ldquo;connect this app to Zapier, Zapier to that API.&amp;rdquo; It works — until it doesn&amp;rsquo;t, and you&amp;rsquo;re debugging a chain of three services for a one-line API call. &lt;a href="https://www.youtube.com/watch?v=kqdsH2NMzyo"&gt;This video&lt;/a&gt; is the counter-argument: &lt;strong&gt;calling OpenAI&amp;rsquo;s Vision API directly from Typebot, no middleman at all.&lt;/strong&gt;&lt;/p&gt;
&lt;h2 id="what-the-build-does"&gt;What the build does&lt;/h2&gt;
&lt;p&gt;The flow: a user uploads an image in the chat, Typebot sends it straight to OpenAI&amp;rsquo;s Vision API (GPT-4V) via an HTTP request block, and the model&amp;rsquo;s description lands back in the conversation. Ask &amp;ldquo;what&amp;rsquo;s in this picture?&amp;rdquo; — get a real answer, all inside one tool.&lt;/p&gt;</description></item><item><title>Charging for Chat: The Stripe Pay-to-Chat Pattern</title><link>https://neferlopez.com/reviews/stripe-pay-to-chat/</link><pubDate>Wed, 18 Oct 2023 00:00:00 +0000</pubDate><guid>https://neferlopez.com/reviews/stripe-pay-to-chat/</guid><description>&lt;p&gt;This is the most-viewed tutorial on the channel for a reason: &lt;a href="https://www.youtube.com/watch?v=Flr-D7xOpfY"&gt;it&amp;rsquo;s the one where the chatbot starts making money&lt;/a&gt;.&lt;/p&gt;
&lt;p&gt;The pitch is simple — &lt;strong&gt;let users chat for free up to a limit, then pay to continue&lt;/strong&gt; — and the video walks the high-level wiring using Typebot, Retune, and Stripe Checkout.&lt;/p&gt;
&lt;h2 id="the-pattern"&gt;The pattern&lt;/h2&gt;
&lt;ol&gt;
&lt;li&gt;&lt;strong&gt;Free questions first.&lt;/strong&gt; The bot answers normally for the first few messages. This is the demo that sells itself — the visitor gets real value before being asked for a cent.&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;The gate.&lt;/strong&gt; Hit the question limit and the flow pivots: here&amp;rsquo;s what you get, here&amp;rsquo;s the price, here&amp;rsquo;s a Stripe Checkout link.&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Payment unlocks the session.&lt;/strong&gt; After checkout, the flow continues into the paid experience — deeper answers, unlimited questions, whatever you promised.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;No middleware servers, no custom auth stack. Typebot holds the conversation logic, Retune provides the AI, Stripe does what Stripe does.&lt;/p&gt;</description></item><item><title>Ask My Books: RAG With Retune's API Inside Typebot</title><link>https://neferlopez.com/reviews/retune-api-ask-your-books/</link><pubDate>Sat, 23 Sep 2023 00:00:00 +0000</pubDate><guid>https://neferlopez.com/reviews/retune-api-ask-your-books/</guid><description>&lt;p&gt;At 2,600+ views, &lt;a href="https://www.youtube.com/watch?v=VIRLbpxGzLU"&gt;this is the most-watched video on the channel&lt;/a&gt; — and I think it&amp;rsquo;s because it answers the question everyone actually has: &lt;strong&gt;&amp;ldquo;Can I make a bot that knows my documents?&amp;rdquo;&lt;/strong&gt; The answer is yes, in an afternoon, with no custom backend.&lt;/p&gt;
&lt;h2 id="the-build"&gt;The build&lt;/h2&gt;
&lt;p&gt;Two moving parts:&lt;/p&gt;
&lt;ol&gt;
&lt;li&gt;&lt;strong&gt;Retune&lt;/strong&gt; gets the brains. Upload the two books as data sources and it builds a searchable knowledge base — this is retrieval-augmented generation (RAG) without you writing any of the retrieval code.&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Typebot&lt;/strong&gt; gets the face. An API call block hits Retune&amp;rsquo;s endpoint, and the user&amp;rsquo;s question comes back as an answer grounded in the books.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;That&amp;rsquo;s the whole architecture. The video shows the API call being configured end to end — endpoint, payload, and mapping the response back into the chat.&lt;/p&gt;</description></item></channel></rss>