{"id":2895,"date":"2024-02-18T15:29:57","date_gmt":"2024-02-18T15:29:57","guid":{"rendered":"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/"},"modified":"2024-02-18T15:29:57","modified_gmt":"2024-02-18T15:29:57","slug":"building-predictive-models-to-improve-debt-collection-process","status":"publish","type":"resource","link":"https:\/\/esisoc.com\/pt\/resource\/criacao-de-modelos-preditivos-para-melhorar-o-processo-de-cobranca-de-dividas\/","title":{"rendered":"Cria\u00e7\u00e3o de modelos preditivos para melhorar o processo de cobran\u00e7a de d\u00edvidas"},"content":{"rendered":"<h2 style=\"text-align: center;\">Principais pormenores<\/h2>\n<p>Aumento de 2 vezes das receitas devido a uma melhor segmenta\u00e7\u00e3o dos clientes.<\/p>\n<div>\n<ul>\n<li>\n<div>Desafio<\/div>\n<div>Melhorar a efic\u00e1cia da cobran\u00e7a de d\u00edvidas com a ajuda da an\u00e1lise preditiva<\/div>\n<\/li>\n<li>\n<div>Solu\u00e7\u00e3o<\/div>\n<div> Um modelo de aprendizagem autom\u00e1tica para prever a probabilidade de promessa de pagamento<\/div>\n<\/li>\n<li>\n<div>Tecnologias e ferramentas<\/div>\n<div>Ecossistema de an\u00e1lise de dados Python, VPN Checkpoint, SQL Server, pacote Lightgbm<\/div>\n<\/li>\n<\/ul>\n<\/div>\n<h2 style=\"text-align: center;\">Cliente<\/h2>\n<p>O cliente \u00e9 uma ag\u00eancia de cobran\u00e7a de d\u00edvidas que cobra d\u00edvidas em v\u00e1rios sectores e clientes. Os principais clientes da ag\u00eancia s\u00e3o os bancos, <a href=\"https:\/\/essidsolutions.com\/industry\/ai-solutions-retail\">retalho<\/a>, empresas de telecomunica\u00e7\u00f5es, empresas p\u00fablicas.<\/p>\n<h2 style=\"text-align: center;\">Desafio: melhorar a efic\u00e1cia da cobran\u00e7a de d\u00edvidas com a ajuda da an\u00e1lise preditiva<\/h2>\n<p>Mais de 1500 agentes de cobran\u00e7as em todo o pa\u00eds lidam com cerca de 3,5 milh\u00f5es de devedores por m\u00eas, contactando mensalmente cerca de 2 milh\u00f5es de devedores.<\/p>\n<p>O processo de cobran\u00e7a de d\u00edvidas inclui as seguintes etapas:<\/p>\n<ol>\n<li>ligar-se a uma conta<\/li>\n<li>verifica\u00e7\u00e3o da conta<\/li>\n<li>promessa de pagamento<\/li>\n<li>cole\u00e7\u00e3o<\/li>\n<\/ol>\n<p>Em conjunto com o respons\u00e1vel pela ci\u00eancia dos dados da empresa, cujo departamento j\u00e1 tinha iniciado a implementa\u00e7\u00e3o da aprendizagem autom\u00e1tica para melhorar a tomada de decis\u00f5es ao longo do ciclo de vida das colec\u00e7\u00f5es, foi decidido que <a href=\"https:\/\/essidsolutions.com\/\">Solu\u00e7\u00f5es ESSID<\/a> explorar\u00e1 o potencial de <a href=\"http:\/\/localhost\/essidsolutions\/service\/predictive-analytics\">an\u00e1lise preditiva<\/a> para identificar os clientes com maior probabilidade de reembolso.<\/p>\n<p>A condi\u00e7\u00e3o indispens\u00e1vel do trabalho era permitir a execu\u00e7\u00e3o de previs\u00f5es na infraestrutura MS SQL existente do cliente.<\/p>\n<h2 style=\"text-align: center;\">Solu\u00e7\u00e3o: modelo de aprendizagem autom\u00e1tica para prever a probabilidade de promessa de pagamento<\/h2>\n<p>A ESSID Solutions come\u00e7ou a trabalhar numa <a href=\"http:\/\/localhost\/essidsolutions\/service\/machine-learning-consulting\">modelo de aprendizagem autom\u00e1tica<\/a> para prever a probabilidade de promessa de pagamento das contas verificadas. Previs\u00f5es exactas devem conduzir a um direcionamento mais priorit\u00e1rio das contas e, consequentemente, a melhores taxas de cobran\u00e7a e custos reduzidos.<\/p>\n<p>O desenvolvimento do modelo preditivo incluiu algumas etapas importantes, como a constru\u00e7\u00e3o de um pipeline para o processamento de dados e a cria\u00e7\u00e3o de carater\u00edsticas no SQL Server, o treino do modelo preditivo com base no lightgbm, a constru\u00e7\u00e3o de um pipeline para obter previs\u00f5es.<\/p>\n<p>A equipa de um engenheiro de dados e de um cientista de dados foi afetada ao projeto, que consistiu nas seguintes fases<\/p>\n<table style=\"height: 438px;\" width=\"711\">\n<tbody>\n<tr>\n<td>\n<p style=\"text-align: left;\"><strong>Est\u00e1gio<\/strong><\/p>\n<\/td>\n<td>\n<p style=\"text-align: left;\"><strong>\u00c2mbito dos trabalhos <\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: left;\">1. Prepara\u00e7\u00e3o dos dados<\/td>\n<td style=\"text-align: left;\">\n<p style=\"text-align: left;\">An\u00e1lise de dados<\/p>\n<p>Limpeza de dados<\/p>\n<p>Cria\u00e7\u00e3o de um pipeline de dados para processamento\/agrega\u00e7\u00e3o de dados<\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: left;\">2. Modela\u00e7\u00e3o<\/td>\n<td>\n<p style=\"text-align: left;\">Desenvolvimento e teste de modelos<\/p>\n<\/td>\n<\/tr>\n<tr>\n<td style=\"text-align: left;\">3. Implanta\u00e7\u00e3o<\/td>\n<td>\n<p style=\"text-align: left;\">Implementa\u00e7\u00e3o no MS SQL 2017, testes de integra\u00e7\u00e3o<\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><span style=\"letter-spacing: 0.6px; -webkit-text-stroke-color: transparent;\">Como resultado do projeto, fornecemos os seguintes resultados ao Cliente:<\/span><\/p>\n<ul>\n<li>M\u00f3dulo Python implement\u00e1vel com:<br \/> - Motor de processamento de dados<br \/> - Motor preditivo<\/li>\n<li>M\u00f3dulo Python implementado no MS SQL 2017:<\/li>\n<li>C\u00f3digo fonte e documenta\u00e7\u00e3o do projeto.<\/li>\n<\/ul>\n<h2 style=\"text-align: center;\">Resultado: maior efic\u00e1cia do processo de cobran\u00e7a de d\u00edvidas<\/h2>\n<p>O modelo preditivo fornecido pela ESSID Solutions prev\u00ea com precis\u00e3o a probabilidade de promessa de pagamento de uma conta.<\/p>\n<p>O desempenho do modelo foi medido pela pontua\u00e7\u00e3o ROC_AUC. A pontua\u00e7\u00e3o ROC_AUC atingiu \u22480,775, o que representou uma melhoria significativa para o Cliente.<\/p>\n<p>Isto d\u00e1 ao Cliente a capacidade de otimizar o tempo dos agentes de cobran\u00e7as, permitindo-lhes visar primeiro as contas mais promissoras.<\/p>","protected":false},"excerpt":{"rendered":"<p>Principais pormenores Aumento de 2 vezes das receitas devido a uma melhor segmenta\u00e7\u00e3o dos clientes. Desafio Melhorar a efic\u00e1cia da cobran\u00e7a de d\u00edvidas com a ajuda da an\u00e1lise preditiva Solu\u00e7\u00e3o Um modelo de aprendizagem autom\u00e1tica para prever a probabilidade de promessa de pagamento Tecnologias e ferramentas Ecossistema de an\u00e1lise de dados Python, VPN Checkpoint, SQL Server, pacote Lightgbm Cliente O cliente \u00e9 uma empresa de cobran\u00e7a de d\u00edvidas ... Ler mais <a title=\"Cria\u00e7\u00e3o de modelos preditivos para melhorar o processo de cobran\u00e7a de d\u00edvidas\" class=\"read-more\" href=\"https:\/\/esisoc.com\/pt\/resource\/criacao-de-modelos-preditivos-para-melhorar-o-processo-de-cobranca-de-dividas\/\" aria-label=\"Leia mais sobre Building Predictive Models to Improve Debt Collection Process\">Ler mais<\/a><\/p>","protected":false},"featured_media":2896,"template":"","industry":[73],"expertise":[93,42,43],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v21.9 (Yoast SEO v21.9.1) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Building Predictive Models to Improve Debt Collection Process - ESISOC | ESSID Solutions<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/esisoc.com\/pt\/resource\/criacao-de-modelos-preditivos-para-melhorar-o-processo-de-cobranca-de-dividas\/\" \/>\n<meta property=\"og:locale\" content=\"pt_PT\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Building Predictive Models to Improve Debt Collection Process\" \/>\n<meta property=\"og:description\" content=\"Key Details Increased revenue 2x times due to improved customer segmentation. Challenge Improve debt collection effectiveness with the help of predictive analytics Solution A machine learning model for predicting the probability of promise to pay Technologies and tools Python data analysis ecosystem, VPN Checkpoint, SQL Server, Lightgbm package Client The Client is a debt collection ... Ler mais\" \/>\n<meta property=\"og:url\" content=\"https:\/\/esisoc.com\/pt\/resource\/criacao-de-modelos-preditivos-para-melhorar-o-processo-de-cobranca-de-dividas\/\" \/>\n<meta property=\"og:site_name\" content=\"ESISOC | ESSID Solutions\" \/>\n<meta property=\"og:image\" content=\"https:\/\/esisoc.com\/wp-content\/uploads\/2024\/02\/6d605982795143dd9537f6580c23b324.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"839\" \/>\n\t<meta property=\"og:image:height\" content=\"514\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Tempo estimado de leitura\" \/>\n\t<meta name=\"twitter:data1\" content=\"2 minutos\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/\",\"url\":\"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/\",\"name\":\"Building Predictive Models to Improve Debt Collection Process - ESISOC | ESSID Solutions\",\"isPartOf\":{\"@id\":\"https:\/\/esisoc.com\/#website\"},\"datePublished\":\"2024-02-18T15:29:57+00:00\",\"dateModified\":\"2024-02-18T15:29:57+00:00\",\"breadcrumb\":{\"@id\":\"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/#breadcrumb\"},\"inLanguage\":\"pt-PT\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/esisoc.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Building Predictive Models to Improve Debt Collection Process\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/esisoc.com\/#website\",\"url\":\"https:\/\/esisoc.com\/\",\"name\":\"ESISOC | ESSID Solutions\",\"description\":\"Data Science Consulting and AI | Online Books, Videos, Courses and more\",\"publisher\":{\"@id\":\"https:\/\/esisoc.com\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/esisoc.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"pt-PT\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/esisoc.com\/#organization\",\"name\":\"ESISOC | ESSID Solutions\",\"url\":\"https:\/\/esisoc.com\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"pt-PT\",\"@id\":\"https:\/\/esisoc.com\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/esisoc.com\/wp-content\/uploads\/2024\/02\/logo-esisoc.png\",\"contentUrl\":\"https:\/\/esisoc.com\/wp-content\/uploads\/2024\/02\/logo-esisoc.png\",\"width\":350,\"height\":63,\"caption\":\"ESISOC | ESSID Solutions\"},\"image\":{\"@id\":\"https:\/\/esisoc.com\/#\/schema\/logo\/image\/\"}}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Building Predictive Models to Improve Debt Collection Process - ESISOC | ESSID Solutions","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/esisoc.com\/pt\/resource\/criacao-de-modelos-preditivos-para-melhorar-o-processo-de-cobranca-de-dividas\/","og_locale":"pt_PT","og_type":"article","og_title":"Building Predictive Models to Improve Debt Collection Process","og_description":"Key Details Increased revenue 2x times due to improved customer segmentation. Challenge Improve debt collection effectiveness with the help of predictive analytics Solution A machine learning model for predicting the probability of promise to pay Technologies and tools Python data analysis ecosystem, VPN Checkpoint, SQL Server, Lightgbm package Client The Client is a debt collection ... Ler mais","og_url":"https:\/\/esisoc.com\/pt\/resource\/criacao-de-modelos-preditivos-para-melhorar-o-processo-de-cobranca-de-dividas\/","og_site_name":"ESISOC | ESSID Solutions","og_image":[{"width":839,"height":514,"url":"https:\/\/esisoc.com\/wp-content\/uploads\/2024\/02\/6d605982795143dd9537f6580c23b324.webp","type":"image\/webp"}],"twitter_card":"summary_large_image","twitter_misc":{"Tempo estimado de leitura":"2 minutos"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/","url":"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/","name":"Building Predictive Models to Improve Debt Collection Process - ESISOC | ESSID Solutions","isPartOf":{"@id":"https:\/\/esisoc.com\/#website"},"datePublished":"2024-02-18T15:29:57+00:00","dateModified":"2024-02-18T15:29:57+00:00","breadcrumb":{"@id":"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/#breadcrumb"},"inLanguage":"pt-PT","potentialAction":[{"@type":"ReadAction","target":["https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/esisoc.com\/resource\/building-predictive-models-to-improve-debt-collection-process\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/esisoc.com\/"},{"@type":"ListItem","position":2,"name":"Building Predictive Models to Improve Debt Collection Process"}]},{"@type":"WebSite","@id":"https:\/\/esisoc.com\/#website","url":"https:\/\/esisoc.com\/","name":"ESISOC | ESSID Solutions","description":"Data Science Consulting and AI | Online Books, Videos, Courses and more","publisher":{"@id":"https:\/\/esisoc.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/esisoc.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"pt-PT"},{"@type":"Organization","@id":"https:\/\/esisoc.com\/#organization","name":"ESISOC | ESSID Solutions","url":"https:\/\/esisoc.com\/","logo":{"@type":"ImageObject","inLanguage":"pt-PT","@id":"https:\/\/esisoc.com\/#\/schema\/logo\/image\/","url":"https:\/\/esisoc.com\/wp-content\/uploads\/2024\/02\/logo-esisoc.png","contentUrl":"https:\/\/esisoc.com\/wp-content\/uploads\/2024\/02\/logo-esisoc.png","width":350,"height":63,"caption":"ESISOC | ESSID Solutions"},"image":{"@id":"https:\/\/esisoc.com\/#\/schema\/logo\/image\/"}}]}},"_links":{"self":[{"href":"https:\/\/esisoc.com\/pt\/wp-json\/wp\/v2\/resource\/2895"}],"collection":[{"href":"https:\/\/esisoc.com\/pt\/wp-json\/wp\/v2\/resource"}],"about":[{"href":"https:\/\/esisoc.com\/pt\/wp-json\/wp\/v2\/types\/resource"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/esisoc.com\/pt\/wp-json\/wp\/v2\/media\/2896"}],"wp:attachment":[{"href":"https:\/\/esisoc.com\/pt\/wp-json\/wp\/v2\/media?parent=2895"}],"wp:term":[{"taxonomy":"industry","embeddable":true,"href":"https:\/\/esisoc.com\/pt\/wp-json\/wp\/v2\/industry?post=2895"},{"taxonomy":"expertise","embeddable":true,"href":"https:\/\/esisoc.com\/pt\/wp-json\/wp\/v2\/expertise?post=2895"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}