{"id":38086,"date":"2026-10-11T18:27:05","date_gmt":"2026-10-11T18:27:05","guid":{"rendered":"https:\/\/thebluemanakin.com\/ai\/ai-agents\/"},"modified":"2026-10-11T18:41:45","modified_gmt":"2026-10-11T18:41:45","slug":"ai-agents","status":"publish","type":"page","link":"https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/","title":{"rendered":"Custom AI agents for companies"},"content":{"rendered":"<h2>An agent is not a chatbot<\/h2>\n<p>A chatbot answers questions. An agent does the work: it reads your CRM, drafts the proposal, files it where it belongs and tells the person who has to sign it. The difference isn&#8217;t the model underneath \u2014 it&#8217;s that the agent has access to your tools and permission to act in them.<\/p>\n<figure class=\"bmb bmb-figure\">\n<div class=\"bmb-scroll\">\n<svg class=\"bmb-svg\" viewBox=\"0 0 760 300\" role=\"img\" aria-labelledby=\"ag-t ag-c\">\n<title id=\"ag-t\">A chatbot answers a question; an agent reads your CRM, drafts, waits for approval, files and notifies<\/title>\n<defs><marker id=\"ag-a\" viewBox=\"0 0 10 10\" refX=\"9\" refY=\"5\" markerWidth=\"7\" markerHeight=\"7\" orient=\"auto-start-reverse\"><path d=\"M0 0L10 5L0 10z\" fill=\"#001FFE\"\/><\/marker><\/defs>\n<text class=\"t-h\" x=\"10\" y=\"30\">Chatbot<\/text>\n<rect x=\"10\" y=\"46\" width=\"170\" height=\"54\" rx=\"16\" fill=\"#F2F4FF\" stroke=\"#CCD2FF\"\/><text class=\"t-lbl\" x=\"95.0\" y=\"78.0\" text-anchor=\"middle\">Question<\/text>\n<path d=\"M182 73L236 73\" stroke=\"#001FFE\" stroke-width=\"2\" fill=\"none\" marker-end=\"url(#ag-a)\"\/>\n<rect x=\"240\" y=\"46\" width=\"170\" height=\"54\" rx=\"16\" fill=\"#F2F4FF\" stroke=\"#CCD2FF\"\/><text class=\"t-lbl\" x=\"325.0\" y=\"78.0\" text-anchor=\"middle\">Answer<\/text>\n<text class=\"t-sm\" x=\"430\" y=\"78\">It answers and stops there.<\/text>\n<line x1=\"10\" y1=\"130\" x2=\"750\" y2=\"130\" stroke=\"#CCD2FF\"\/>\n<text class=\"t-h\" x=\"10\" y=\"172\">Agent<\/text>\n<rect x=\"10\" y=\"188\" width=\"160\" height=\"60\" rx=\"16\" fill=\"#F2F4FF\" stroke=\"#CCD2FF\"\/><text class=\"t-lbl\" x=\"90.0\" y=\"223.0\" text-anchor=\"middle\">Reads your CRM<\/text>\n<path d=\"M172 218L199 218\" stroke=\"#001FFE\" stroke-width=\"2\" fill=\"none\" marker-end=\"url(#ag-a)\"\/>\n<rect x=\"203\" y=\"188\" width=\"160\" height=\"60\" rx=\"16\" fill=\"#001FFE\" stroke=\"#001FFE\"\/><text class=\"t-lbl t-on\" x=\"283.0\" y=\"223.0\" text-anchor=\"middle\">Drafts the proposal<\/text>\n<path d=\"M365 218L392 218\" stroke=\"#001FFE\" stroke-width=\"2\" fill=\"none\" marker-end=\"url(#ag-a)\"\/>\n<rect x=\"396\" y=\"188\" width=\"160\" height=\"60\" rx=\"16\" fill=\"#F2F4FF\" stroke=\"#CCD2FF\"\/><text class=\"t-lbl\" x=\"476.0\" y=\"223.0\" text-anchor=\"middle\">Files it in place<\/text>\n<path d=\"M558 218L585 218\" stroke=\"#001FFE\" stroke-width=\"2\" fill=\"none\" marker-end=\"url(#ag-a)\"\/>\n<rect x=\"589\" y=\"188\" width=\"160\" height=\"60\" rx=\"16\" fill=\"#F2F4FF\" stroke=\"#CCD2FF\"\/><text class=\"t-lbl\" x=\"669.0\" y=\"223.0\" text-anchor=\"middle\">Notifies the signer<\/text>\n<rect x=\"333\" y=\"262\" width=\"126\" height=\"26\" rx=\"13\" fill=\"#fff\" stroke=\"#001FFE\" stroke-dasharray=\"4 3\"\/>\n<text class=\"t-sm\" x=\"396\" y=\"279\" text-anchor=\"middle\">Human approval<\/text>\n<path d=\"M396 262V222\" stroke=\"#001FFE\" stroke-width=\"1.5\" stroke-dasharray=\"3 3\"\/>\n<\/svg><\/div><figcaption id=\"ag-c\">The difference is not the model: it is access to your tools and permission to act in them.<\/figcaption><\/figure>\n<p>That is also why most corporate AI projects never reach production. A licence gets bought, a chat window opens, and everyone is left to invent a use for it. With no access to real data and no specific job to do, the result is one more tool that nobody opens after week three.<\/p>\n<h2>What we do<\/h2>\n<ul>\n<li><strong>Agents wired into your systems.<\/strong> We build agents on Claude, GPT and whichever models suit each task, connected over MCP or API to your CRM, your inbox, your document store, your spreadsheet or your database. The agent doesn&#8217;t &#8220;know&#8221; your company: it reads your data when it needs it.<\/li>\n<li><strong>One specific job, not a generic tool.<\/strong> We start from a task someone does by hand today and that fits in a sentence: qualify inbound leads, prepare the weekly report, answer the first email of a quote, check invoices against the contract.<\/li>\n<li><strong>Human control where it matters.<\/strong> We decide with you what the agent does alone and what goes past a person before it leaves the building. An agent that drafts and waits for approval gets adopted; one that sends without asking gets switched off after the first scare.<\/li>\n<li><strong>Evaluation before trust.<\/strong> We build a set of real cases with known correct answers and measure the agent against them. Without that test there is no way to know whether it works \u2014 only the impression of whoever ran the demo.<\/li>\n<li><strong>Documented handover.<\/strong> The agent stays in your house, with its instructions, its permissions and its limits written in plain language, so your team can change it without calling us.<\/li>\n<\/ul>\n<h2>How it starts<\/h2>\n<p>Before building anything we look at which tasks your team repeats, how many hours go into them and which ones have a verifiable answer. That last part is what decides: a task whose output nobody can score cannot be automated safely, however repetitive it is.<\/p>\n<p>Out of that comes a short list ranked by hours saved and by risk. We start with one, put it into production and measure it for a few weeks before touching the next.<\/p>\n<h2>What we measure<\/h2>\n<p>We agree the number before we start. Depending on the case it is usually hours freed per month, response time to the customer, share of tasks the agent closes without a human, or errors caught that used to slip through.<\/p>\n<p>We don&#8217;t report &#8220;number of model calls&#8221; or cost per token as if they were results. They are costs, we watch them, but they are not what you bought.<\/p>\n<h2>Why us<\/h2>\n<p>We use agents in our own work every day: in campaign data analysis, in content production and in maintaining the sites we run. What we propose is what already works inside the agency, not a service line invented for the brochure.<\/p>\n<p>We also know where it breaks. An agent with access to your systems can do damage if it is given broad permissions and allowed to read text that comes from outside: an email or a document can carry instructions aimed at the model. So we separate reading from writing, scope permissions per task, and log everything the agent does.<\/p>\n<h2>The rest of the artificial intelligence area<\/h2>\n<ul class=\"bm-hermanas\">\n<li><a href=\"https:\/\/thebluemanakin.com\/en\/ai\/ai-process-automation\/\">Internal process automation<\/a><\/li>\n<li><a href=\"https:\/\/thebluemanakin.com\/en\/ai\/ai-search-visibility\/\">Visibility in AI search and assistants<\/a><\/li>\n<li><a href=\"https:\/\/thebluemanakin.com\/en\/ai\/ai-content-at-scale\/\">Content at scale under editorial control<\/a><\/li>\n<li><a href=\"https:\/\/thebluemanakin.com\/en\/ai\/\">Artificial intelligence: overview<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>An agent is not a chatbot A chatbot answers questions. An agent does the work: it reads your CRM, drafts the proposal, files it where it belongs and tells the person who has to sign it. The difference isn&#8217;t the model underneath \u2014 it&#8217;s that the agent has access to your tools and permission to act in them. A chatbot answers a question; an agent reads your CRM, drafts, waits for approval, files and notifies Chatbot Question Answer It answers and stops there. Agent Reads your CRM Drafts the proposal Files it in place Notifies the signer Human approval The difference is not the model: it is access to your tools and permission to act in them. That is also why most corporate AI projects never reach production. A licence gets bought, a chat window opens, and everyone is left to invent a use for it. With no access to real data and no specific job to do, the result is one more tool that nobody opens after week three. What we do Agents wired into your systems. We build agents on Claude, GPT and whichever models suit each task, connected over MCP or API to your CRM, your inbox, your document store, your spreadsheet or your database. The agent doesn&#8217;t &#8220;know&#8221; your company: it reads your data when it needs it. One specific job, not a generic tool. We start from a task someone does by hand today and that fits in a sentence: qualify inbound leads, prepare the weekly report, answer the first email of a quote, check invoices against the contract. Human control where it matters. We decide with you what the agent does alone and what goes past a person before it leaves the building. An agent that drafts and waits for approval gets adopted; one that sends without asking gets switched off after the first scare. Evaluation before trust. We build a set of real cases with known correct answers and measure the agent against them. Without that test there is no way to know whether it works \u2014 only the impression of whoever ran the demo. Documented handover. The agent stays in your house, with its instructions, its permissions and its limits written in plain language, so your team can change it without calling us. How it starts Before building anything we look at which tasks your team repeats, how many hours go into them and which ones have a verifiable answer. That last part is what decides: a task whose output nobody can score cannot be automated safely, however repetitive it is. Out of that comes a short list ranked by hours saved and by risk. We start with one, put it into production and measure it for a few weeks before touching the next. What we measure We agree the number before we start. Depending on the case it is usually hours freed per month, response time to the customer, share of tasks the agent closes without a human, or errors caught that used to slip through. We don&#8217;t report &#8220;number of model calls&#8221; or cost per token as if they were results. They are costs, we watch them, but they are not what you bought. Why us We use agents in our own work every day: in campaign data analysis, in content production and in maintaining the sites we run. What we propose is what already works inside the agency, not a service line invented for the brochure. We also know where it breaks. An agent with access to your systems can do damage if it is given broad permissions and allowed to read text that comes from outside: an email or a document can carry instructions aimed at the model. So we separate reading from writing, scope permissions per task, and log everything the agent does. The rest of the artificial intelligence area Internal process automation Visibility in AI search and assistants Content at scale under editorial control Artificial intelligence: overview<\/p>\n","protected":false},"author":6,"featured_media":0,"parent":31211,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"page-bm-servicio.php","meta":{"content-type":"","telegram_tosend":false,"telegram_tosend_message":"","telegram_tosend_target":0,"footnotes":""},"class_list":["post-38086","page","type-page","status-publish","hentry"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.5 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Custom AI agents for companies | Blue Manakin<\/title>\n<meta name=\"description\" content=\"Agents on Claude and GPT connected to your CRM, inbox and documents. One specific job, human control, and hours freed as the measure.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Custom AI agents for companies | Blue Manakin\" \/>\n<meta property=\"og:description\" content=\"Agents on Claude and GPT connected to your CRM, inbox and documents. One specific job, human control, and hours freed as the measure.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/\" \/>\n<meta property=\"og:site_name\" content=\"Blue Manakin\" \/>\n<meta property=\"article:modified_time\" content=\"2026-10-11T18:41:45+00:00\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:site\" content=\"@thebluemanakin\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"3 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/en\\\/ai\\\/ai-agents\\\/\",\"url\":\"https:\\\/\\\/thebluemanakin.com\\\/en\\\/ai\\\/ai-agents\\\/\",\"name\":\"Custom AI agents for companies | Blue Manakin\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/#website\"},\"datePublished\":\"2026-10-11T18:27:05+00:00\",\"dateModified\":\"2026-10-11T18:41:45+00:00\",\"description\":\"Agents on Claude and GPT connected to your CRM, inbox and documents. One specific job, human control, and hours freed as the measure.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/en\\\/ai\\\/ai-agents\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/thebluemanakin.com\\\/en\\\/ai\\\/ai-agents\\\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/en\\\/ai\\\/ai-agents\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/thebluemanakin.com\\\/en\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Artificial Intelligence Agency\",\"item\":\"https:\\\/\\\/thebluemanakin.com\\\/en\\\/ai\\\/\"},{\"@type\":\"ListItem\",\"position\":3,\"name\":\"Custom AI agents for companies\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/#website\",\"url\":\"https:\\\/\\\/thebluemanakin.com\\\/\",\"name\":\"Blue Manakin\",\"description\":\"Blue Manakin. Come Dance With Us!\",\"publisher\":{\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/thebluemanakin.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/#organization\",\"name\":\"Blue Manakin\",\"url\":\"https:\\\/\\\/thebluemanakin.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/thebluemanakin.com\\\/wp-content\\\/uploads\\\/2021\\\/07\\\/cropped-logo-header.png\",\"contentUrl\":\"https:\\\/\\\/thebluemanakin.com\\\/wp-content\\\/uploads\\\/2021\\\/07\\\/cropped-logo-header.png\",\"width\":613,\"height\":250,\"caption\":\"Blue Manakin\"},\"image\":{\"@id\":\"https:\\\/\\\/thebluemanakin.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/thebluemanakin\",\"https:\\\/\\\/www.instagram.com\\\/thebluemanakin\\\/\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/blue-manakin\"]}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Custom AI agents for companies | Blue Manakin","description":"Agents on Claude and GPT connected to your CRM, inbox and documents. One specific job, human control, and hours freed as the measure.","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:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/","og_locale":"en_US","og_type":"article","og_title":"Custom AI agents for companies | Blue Manakin","og_description":"Agents on Claude and GPT connected to your CRM, inbox and documents. One specific job, human control, and hours freed as the measure.","og_url":"https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/","og_site_name":"Blue Manakin","article_modified_time":"2026-10-11T18:41:45+00:00","twitter_card":"summary_large_image","twitter_site":"@thebluemanakin","twitter_misc":{"Est. reading time":"3 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/","url":"https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/","name":"Custom AI agents for companies | Blue Manakin","isPartOf":{"@id":"https:\/\/thebluemanakin.com\/#website"},"datePublished":"2026-10-11T18:27:05+00:00","dateModified":"2026-10-11T18:41:45+00:00","description":"Agents on Claude and GPT connected to your CRM, inbox and documents. One specific job, human control, and hours freed as the measure.","breadcrumb":{"@id":"https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/thebluemanakin.com\/en\/ai\/ai-agents\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/thebluemanakin.com\/en\/"},{"@type":"ListItem","position":2,"name":"Artificial Intelligence Agency","item":"https:\/\/thebluemanakin.com\/en\/ai\/"},{"@type":"ListItem","position":3,"name":"Custom AI agents for companies"}]},{"@type":"WebSite","@id":"https:\/\/thebluemanakin.com\/#website","url":"https:\/\/thebluemanakin.com\/","name":"Blue Manakin","description":"Blue Manakin. Come Dance With Us!","publisher":{"@id":"https:\/\/thebluemanakin.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/thebluemanakin.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/thebluemanakin.com\/#organization","name":"Blue Manakin","url":"https:\/\/thebluemanakin.com\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/thebluemanakin.com\/#\/schema\/logo\/image\/","url":"https:\/\/thebluemanakin.com\/wp-content\/uploads\/2021\/07\/cropped-logo-header.png","contentUrl":"https:\/\/thebluemanakin.com\/wp-content\/uploads\/2021\/07\/cropped-logo-header.png","width":613,"height":250,"caption":"Blue Manakin"},"image":{"@id":"https:\/\/thebluemanakin.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/thebluemanakin","https:\/\/www.instagram.com\/thebluemanakin\/","https:\/\/www.linkedin.com\/company\/blue-manakin"]}]}},"_links":{"self":[{"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/pages\/38086","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/comments?post=38086"}],"version-history":[{"count":2,"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/pages\/38086\/revisions"}],"predecessor-version":[{"id":38134,"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/pages\/38086\/revisions\/38134"}],"up":[{"embeddable":true,"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/pages\/31211"}],"wp:attachment":[{"href":"https:\/\/thebluemanakin.com\/en\/wp-json\/wp\/v2\/media?parent=38086"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}