Laravel AI Preflight Checker with Jev Probability Scores | Mohamed Said [Skip to content](#main) [ ![](https://cdn.msaied.com/01KT78WE565VEMM3PSNQAAB0MH.png) Mohamed SaidLaravel Backend Engineer ](https://www.msaied.com/public) - [Home](https://www.msaied.com/public) - [Projects](https://www.msaied.com/public/projects) - [Articles](https://www.msaied.com/public/articles) - [Certificates](https://www.msaied.com/public/certificates) - [About](https://www.msaied.com/public#about) [ Contact](https://www.msaied.com/public#contact) Menu Menu ---- Close - [HomeStart here](https://www.msaied.com/public) - [ProjectsCase studies](https://www.msaied.com/public/projects) - [ArticlesEngineering notes](https://www.msaied.com/public/articles) - [CertificatesCredentials](https://www.msaied.com/public/certificates) - [AboutHow I work](https://www.msaied.com/public#about) - [ContactGet in touch](https://www.msaied.com/public#contact) [Start a conversation](https://www.msaied.com/public#contact) [WhatsApp](https://wa.me/201094619204) [Email](mailto:hello@msaied.com) 1. [Home](https://www.msaied.com/public) 2. / 3. [Articles](https://www.msaied.com/public/articles) 4. / 5. [Laravel](https://www.msaied.com/public/articles?category=laravel) 6. / 7. Decide with Jev: Build a Laravel AI Content Preflight Checker That Returns a Probability [Laravel](https://www.msaied.com/public/articles?category=laravel) [AI](https://www.msaied.com/public/articles?category=ai) Decide with Jev: Build a Laravel AI Content Preflight Checker That Returns a Probability ========================================================================================= Jev is a TypeSafe AI model that returns a probability score instead of text. Learn how Harris Raftopoulos uses it with the Laravel AI SDK to build a content preflight checker that flags draft/brief mismatches before they reach an editor. ![](https://cdn.msaied.com/01M22N44A70A5MC2S599JP0MPH.webp) [Mohamed Said](https://www.msaied.com/public#person) Published 25 Sep 2026 · Updated 25 Sep 2026 · 4 min read ShareCopy linkCopied ![Decide with Jev: Build a Laravel AI Content Preflight Checker That Returns a Probability](https://cdn.msaied.com/701/916a0dd2b427c6335395d6d2684524ab.png) On this page +1. [What Is "Decide with Jev"?](#what-is-quotdecide-with-jevquot) 2. [Why a Probability Score Instead of Text?](#why-a-probability-score-instead-of-text) 3. [Setting Up the Project](#setting-up-the-project) 4. [The Four Test Drafts](#the-four-test-drafts) 5. [First Results in Tinkerwell](#first-results-in-tinkerwell) 6. [What Comes Next](#what-comes-next) 7. [Key Takeaways](#key-takeaways) What Is "Decide with Jev"? -------------------------- [Decide with Jev](https://laravel-news.com/decide-with-jev-laravel-ai-that-answers-with-a-probability) is a new mini course on Laravel News by Harris Raftopoulos. The goal is to build a **content preflight checker** — an app that compares a written draft against its brief and tells you how well the two align, before the draft ever reaches an editor. The analogy is deliberate: pilots run a preflight checklist before takeoff. This app runs the same kind of check on content. Why a Probability Score Instead of Text? ---------------------------------------- Most AI models you interact with return a paragraph. Jev, an AI model from a company called TypeSafe, returns a **number between 0 and 1** — the probability that the answer to your yes/no question is "yes." That distinction matters in PHP: - A paragraph is hard to act on programmatically. - A number is easy to route through business logic. A score of `0.96` means the draft almost certainly satisfies the brief. A score of `0.02` means it almost certainly does not. Anything in the middle can be flagged for human review. PHP can enforce thresholds on a float; it cannot reliably parse intent from a sentence. > **Important caveat:** The number is the model's confidence in "yes" — it is not proof the answer is correct. Setting Up the Project ---------------------- Episode 1 covers installation and first contact with the API. ```bash # Install the Laravel AI SDK (stable branch now includes this feature) composer require laravel/ai ``` Add your TypeSafe API key to `.env`: ```env TYPESAFE_API_KEY=your-key-here ``` Jev runs on TypeSafe's servers, so there is nothing extra to install locally. > **Note:** The video was recorded against the `1.x` dev branch of the Laravel AI SDK because the Jev feature had not yet merged. It has since landed on the main branch, so a standard `composer require` is all you need. The Four Test Drafts -------------------- To have something meaningful to score, Harris wrote four short drafts for a single brief: *show a Laravel developer how to add a `GET /hello` route and verify it in the browser.* | Draft | Description | |---|---| | 1 | Complete — covers every step the brief requires | | 2 | Mentions routes but provides no actionable steps | | 3 | Mixed — some opinion and product news, only one instruction | | 4 | Adversarial — instructs the model to ignore the brief | Draft 4 is reserved for a later episode where prompt-injection resistance is tested. First Results in Tinkerwell --------------------------- With the SDK configured, Harris queries Jev directly from Tinkerwell: ```php use Laravel\AI\Facades\AI; $score = AI::jev()->decide( question: 'Does this draft fulfill the brief?', context: $brief . '\n\n' . $draft ); // Draft 1 (complete) → 0.96 // Draft 2 (thin) → 0.02 ``` The gap between `0.96` and `0.02` is exactly what you want from a classifier: a clear signal at both ends of the scale. What Comes Next --------------- Episode 2 expands the single question into three parallel checks: 1. Does the draft fulfill the brief? 2. What type of content is it? 3. How clear are the instructional steps? Running multiple Jev queries gives you a richer signal without asking the model to produce prose you then have to parse. Key Takeaways ------------- - **Jev returns a probability**, not text — making it directly usable in PHP conditionals and routing logic. - **The Laravel AI SDK** (stable branch) now supports Jev out of the box; no dev branch required. - **A content preflight checker** is a practical, low-risk way to introduce AI into an editorial workflow without removing human oversight. - **Probability ≠ truth** — treat the score as a signal, not a verdict. - The project source is available at [github.com/harris21/decide-with-jev](https://github.com/harris21/decide-with-jev). --- *Source: [Decide with Jev: Laravel AI That Answers with a Probability — Laravel News](https://laravel-news.com/decide-with-jev-laravel-ai-that-answers-with-a-probability)* - [Laravel AI](https://www.msaied.com/public/articles?search=Laravel%20AI) - [Jev](https://www.msaied.com/public/articles?search=Jev) - [TypeSafe](https://www.msaied.com/public/articles?search=TypeSafe) - [AI SDK](https://www.msaied.com/public/articles?search=AI%20SDK) - [content automation](https://www.msaied.com/public/articles?search=content%20automation) - [PHP AI](https://www.msaied.com/public/articles?search=PHP%20AI) Frequently asked questions --------------------------- What makes Jev different from a standard chat AI model in a Laravel app?Instead of returning a text response, Jev returns a probability score between 0 and 1 representing how likely the answer to a yes/no question is "yes". This makes the output directly usable in PHP conditionals without any text parsing. Do I need a special branch of the Laravel AI SDK to use Jev?No. The tutorial was originally recorded against a 1.x dev branch, but the Jev feature has since been merged into the main branch. A standard `composer require laravel/ai` install is sufficient. Can a high Jev probability score be treated as a definitive answer?No. The score reflects the model's confidence that the answer is "yes" — it is not proof the answer is correct. The course recommends routing borderline scores to a human reviewer rather than acting on them automatically. ![Mohamed Said](https://cdn.msaied.com/01M22N44A70A5MC2S599JP0MPH.webp)About the author ---------------- [Mohamed Said](https://www.msaied.com/public#person)Senior Backend Engineer specializing in Laravel, scalable SaaS platforms, APIs, and cloud infrastructure. I build secure, high-performance web applications that help businesses grow. [About](https://www.msaied.com/public#about) [GitHub ↗](https://github.com/EG-Mohamed) [LinkedIn ↗](https://www.linkedin.com/in/msaiedm/) [WhatsApp ↗](https://wa.me/201094619204) [Email Address ↗](mailto:hello@msaied.com) [My CV ↗](https://drive.google.com/file/u/0/d/1MF20IPRJyzfy32mhEutjL5EpSls0w2Q8/view) [Previous articleFilament v4 Migrating from v3: Breaking Changes and Refactor Patterns](https://www.msaied.com/public/articles/filament-v4-migrating-from-v3-breaking-changes-and-refactor-patterns-1) [Next articleProduction AI Agents in Laravel: Streaming, Token Budgets, and Structured Output Contracts](https://www.msaied.com/public/articles/production-ai-agents-in-laravel-streaming-token-budgets-and-structured-output-contracts-4) On this page ------------- 1. [What Is "Decide with Jev"?](#what-is-quotdecide-with-jevquot) 2. [Why a Probability Score Instead of Text?](#why-a-probability-score-instead-of-text) 3. [Setting Up the Project](#setting-up-the-project) 4. [The Four Test Drafts](#the-four-test-drafts) 5. [First Results in Tinkerwell](#first-results-in-tinkerwell) 6. [What Comes Next](#what-comes-next) 7. [Key Takeaways](#key-takeaways) ### Have a technical challenge? Tell me what you’re building. I reply within two working days. [Start a conversation](https://www.msaied.com/public#contact) Related articles ----------------- [ ![](https://cdn.msaied.com/740/cce86edc21eddcbdd2f2454fadaf9c70.png) · 3 min read### The Pipeline Pattern in Laravel: Custom Pipelines Beyond Middleware 5 Oct 2026 ](https://www.msaied.com/public/articles/the-pipeline-pattern-in-laravel-custom-pipelines-beyond-middleware-1) [ ![](https://cdn.msaied.com/739/2d6897fdcdcf090613f96f72a64b8a78.png) · 4 min read### MySQL Full-Text Search in Laravel: Indexes, Relevance Scoring, and Boolean Mode 4 Oct 2026 ](https://www.msaied.com/public/articles/mysql-full-text-search-in-laravel-indexes-relevance-scoring-and-boolean-mode) [ ![](https://cdn.msaied.com/738/073696a3fefe18bec825beec5ac658f5.png) · 4 min read### Laravel Queue Rate-Limited Middleware: Throttling Jobs Without Losing Work 4 Oct 2026 ](https://www.msaied.com/public/articles/laravel-queue-rate-limited-middleware-throttling-jobs-without-losing-work) Have a technical challenge? ---------------------------- Tell me what you’re building. I reply within two working days. [Discuss your project ↗](https://www.msaied.com/public#contact) © 2026 Mohamed Said · Built with Laravel, meant to last.Senior Backend Engineer specializing in Laravel, scalable SaaS platforms, APIs, and cloud infrastructure. I build secure, high-performance web applications that help businesses grow. - [Home](https://www.msaied.com/public) - [Articles](https://www.msaied.com/public/articles) - [Certificates](https://www.msaied.com/public/certificates) - [GitHub](https://github.com/EG-Mohamed) - [LinkedIn](https://www.linkedin.com/in/msaiedm/) - [WhatsApp](https://wa.me/201094619204) - [Email Address](mailto:hello@msaied.com) - [My CV](https://drive.google.com/file/u/0/d/1MF20IPRJyzfy32mhEutjL5EpSls0w2Q8/view) - [Sitemap](https://www.msaied.com/public/sitemap.xml)