--- title: Explainable AI in Production: A Neuro-Symbolic Model for... description: SHAP KernelExplainer takes ~30 ms per prediction and produces stochastic outputs, making it unsuitable for real-time fraud detection. A neuro-symbolic model is... canonical: https://daily.dev/posts/explainable-ai-in-production-a-neuro-symbolic-model-for-real-time-fraud-detection-f4v64el7h twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Explainable AI in Production: A Neuro-Symbolic Model for Real-Time Fraud Detection | daily.dev og:description: SHAP KernelExplainer takes ~30 ms per prediction and produces stochastic outputs, making it unsuitable for real-time fraud detection. A neuro-symbolic model is... og:url: https://daily.dev/posts/explainable-ai-in-production-a-neuro-symbolic-model-for-real-time-fraud-detection-f4v64el7h og:image: https://api.daily.dev/og/posts/f4V64eL7h.png og:image:alt: Explainable AI in Production: A Neuro-Symbolic Model for Real-Time Fraud Detection og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Explainable AI in Production: A Neuro-Symbolic Model for Real-Time Fraud Detection **[Towards Data Science](https://daily.dev/sources/tds)** · 15 min read · 0 upvotes · 0 comments ## Summary SHAP KernelExplainer takes ~30 ms per prediction and produces stochastic outputs, making it unsuitable for real-time fraud detection. A neuro-symbolic model is benchmarked that embeds six differentiable symbolic rules directly into the forward pass, generating deterministic, human-readable explanations in 0.9 ms — a 33× speedup. Tested on the Kaggle Credit Card Fraud dataset, the neuro-symbolic model achieves identical fraud recall (0.8469) with only a minor AUC drop (0.9688 vs 0.9737). The article details the architecture (neural backbone + symbolic rule layer + fusion layer), training results, learned thresholds, and known limitations including a V4 weight collapse where one rule accumulates 57% of symbolic weight. The conclusion is that post-hoc explainability methods are too slow and inconsistent for production fraud systems, and explainability must be baked into the model architecture itself. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [How a Neural Network Learned Its Own Fraud Rules: A Neuro-Symbolic AI Experiment](https://daily.dev/posts/how-a-neural-network-learned-its-own-fraud-rules-a-neuro-symbolic-ai-experiment-a4siik3dn) · Towards Data Science · 1 upvotes · 0 comments - [What SHAP Can't Explain About Agentic AI Fraud](https://daily.dev/posts/what-shap-can-t-explain-about-agentic-ai-fraud-jteoojxfh) · Towards Data Science · 0 upvotes · 0 comments - [I Trained Six Models for Fraud Detection, and the Best One Isn't in Production](https://daily.dev/posts/i-trained-six-models-for-fraud-detection-and-the-best-one-isn-t-in-production-owj2jt66q) · Towards Data Science · 0 upvotes · 0 comments --- Tags: [#pytorch](https://daily.dev/tags/pytorch), [#fraud-detection](https://daily.dev/tags/fraud-detection), [#explainable-ai](https://daily.dev/tags/explainable-ai) [View this post on daily.dev](https://daily.dev/posts/explainable-ai-in-production-a-neuro-symbolic-model-for-real-time-fraud-detection-f4v64el7h) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Explainable AI in Production: A Neuro-Symbolic Model for Real-Time Fraud Detection","url":"https://daily.dev/posts/explainable-ai-in-production-a-neuro-symbolic-model-for-real-time-fraud-detection-f4v64el7h","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/explainable-ai-in-production-a-neuro-symbolic-model-for-real-time-fraud-detection-f4v64el7h"},"datePublished":"2026-03-30T11:24:39.655Z","dateModified":"2026-03-30T11:25:23.199Z","description":"SHAP KernelExplainer takes ~30 ms per prediction and produces stochastic outputs, making it unsuitable for real-time fraud detection. A neuro-symbolic model is...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/38b4bdc1c3b096d79019d5b33d44e427?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/38b4bdc1c3b096d79019d5b33d44e427?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Towards Data Science","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Towards Data Science","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/tds","url":"https://daily.dev/sources/tds"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/explainable-ai-in-production-a-neuro-symbolic-model-for-real-time-fraud-detection-f4v64el7h","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"pytorch,fraud-detection,explainable-ai","timeRequired":"PT15M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Towards Data Science","item":"https://daily.dev/sources/tds"},{"@type":"ListItem","position":3,"name":"Explainable AI in Production: A Neuro-Symbolic Model for Real-Time Fraud Detection"}]} ```