--- title: "StudioChat: Custom AI Agents, Integrated Into Your Stack" description: "We partner with companies to build custom AI agents that plug into the CRM, ERP and WhatsApp they already use. In production: Coderhouse, Takenos, Avera." url: https://studiochat.io locale: en translations: { es: https://studiochat.io/es/index.md, pt: https://studiochat.io/pt/index.md } site: https://studiochat.io --- Conversational AI agents for customers, employees and suppliers # The best salesperson. Everywhere. 24/7. The judgment of the team's best, cloned into one agent for customers, employees, suppliers and partners. [Talk to us now](https://cal.com/studiochat/25min) [See customer stories →](/case-studies) Mica · ACME Supply online End-to-end encrypted TODAY Agent Mica · working Live WhatsApp Collections · Mica Email Post-sale · CamiMobile app Support · Lola Running in production for Coderhouse Takenos Rebill Mudafy Avera Belo Toribio Achával Inarch WannaColive Coderhouse Takenos Rebill Mudafy Avera Belo Toribio Achával Inarch WannaColive TRUSTED BY ## What our clients say +80% autonomous resolution > “StudioChat became our best salesperson: available 24/7, knows every course inside out, and the ROI was immediate.” Christian Patiño Founder & CEO · Coderhouse Coderhouse [Read the story →](/case-studies/coderhouse) 70% autonomous resolution > “Joaquina resolves 70% of conversations on her own, but what we value most is the follow-up they give us as clients and the personalized attention: they adapt to whatever the team needs.” Melina Albornoz CX Team Lead · Takenos Takenos [Read the story →](/case-studies/takenos) 24/7 end-to-end post-sale > “The entire post-sale journey is handled by Eva, our StudioChat assistant, connected to Amazon, Mercado Libre, our official store and WhatsApp. The team now focuses on the cases that truly need them.” Jaime Bravo CX & Post-Sale Operations Manager · Avera Avera [Read the story →](/case-studies/avera) 60% of support with no human intervention > “We automated 60% of what reaches our support inbox. Today I'm the only person dedicated to customer support for hundreds of accounts, and that's possible thanks to Billy. The time that used to go into repeated tickets now goes into the business.” Henry Rangel Head of Customer Success & Support · Rebill Rebill [Read the story →](/case-studies/rebill) HOW IT WORKS ## Assistants that think like the team. And talk like the team. An assistant is built from four things: the knowledge the team already has in whatever format it exists, one skill for each situation it has to resolve, real examples that carry the tone, and connections to the tools the team already works in. ### Knowledge bases in any format. Knowledge bases in whatever format already exists (text, documents, PDFs, images, floor plans), plus the team's internal APIs. The assistant pulls the information it needs, in whatever format it's in, before responding. ### Skills that improve the quality of every response. Every situation (deposits, withdrawals, KYC, account changes) gets handled with a skill trained on the playbook the team already validated. SAMI, StudioChat's own agent, helps improve them over time. ### Examples to clone the tone. Starting from real examples from the best agent, the assistant captures how it opens, how it reassures, how it closes, and replicates that voice in every conversation. ### Connected to the team's tools. Create tickets in Intercom, send Slack alerts, open tasks: the assistant acts on the company's systems in real time, not just replies. NEW Journeys ## Agents that pursue long-horizon goals Beyond the conversation: agents that work one goal for days, weeks or months. They spot the signal, build the plan, remember everything and reach the outcome. [Meet Journeys](/journeys) Now available for the first customers * They act on business signals, not just on incoming messages * Memory that compounds across every touchpoint * Priced per outcome, like all of StudioChat ## Multiple interfaces for the same assistants. Same brain, different channels. The assistant keeps context and a consistent voice in every conversation, whether it's chat, email, or a call. Mobile SDK In-app chat Messaging apps Email Voice We plug into whatever tool is already in use, as one more agent, but a non-human one. Intercom Zendesk WhatsApp Slack HubSpot \+ other tools WhatsApp in production ## Three different processes, one channel. Sales, lead qualification and support: three companies run them on WhatsApp today with a StudioChat agent. The numbers come from those accounts, not from a pilot. Sales Coderhouse logo 5% inquiry-to-sale conversion and over 80% autonomous resolution on the WhatsApp channel. [Message the sales agent](https://stdcht.io/coderhouse-demo) This is Coderhouse's real advisor line. Test it in earnest, and skip the fake enrolment requests. Support Takenos logo 70% autonomous resolution, reading receipts and screenshots inside the conversation. [Message the support agent](https://wa.me/5491124716668) This is live fintech support. Ask it anything you like; just don't open fake cases or force a handoff to a person. Qualification Mudafy logo Hundreds of inquiries a month qualified before an agent steps in, WhatsApp voice notes included. [Read the case study](/case-studies/mudafy) This one isn't a public line. The case study carries the numbers and the deployment detail. There are real support teams working behind the Coderhouse and Takenos numbers. They're open for anyone to test, which is exactly why they aren't a toy. [Everything that runs on WhatsApp](/whatsapp) ## The team defines when to hand off. When the customer asks for it, when a condition the team defined triggers, or when the assistant can't help, the conversation moves to the human team with full context. Explicit customer request Negative sentiment or frustration Skills not covered by the assistant VIP or critical accounts Can I talk to a person? I'd rather sort this out over the phone. IA Of course! I'm handing you to Andrés from the team, who already has your full case. Humano Hi, I'm Andrés 👋 I can see your refund R-2210 is approved. Want me to call you now and close this out? Solo or as part of a team The agent can operate on its own, or as one more member of the human team, handing off exactly when it should. Horizontal or vertical Covers the first steps of several processes (horizontal), or owns one process end to end (vertical), whichever fits. AGENT SECURITY ## Security at every layer, from the model to the client's data. LLM security, agent security, client control, and infrastructure aren't the same layer: each protects something different, and the agent leaves a trail of everything it does along the way. LLM SECURITY: GUARDRAILS ### A protective layer around the model. A language model can make the wrong call: delete something it shouldn't, invent data that doesn't exist (like a bank account number), or let spend spiral out of control. Guardrails validate what the agent is about to answer before it answers, automatically or manually, covering hallucinations and prompt injection. CLIENT CONTROL ### The client decides what information reaches StudioChat. Every integration goes through a middleware layer the client configures, and StudioChat can't see the data the client chooses not to share. Control stays on the client's side, by design. AGENT TRANSPARENCY ### The agent explains what it does, why, and where each fact comes from. The most common fear is that the agent does something and no one finds out. That's why every answer comes with its reasoning and its citations. No black box acting without leaving a trace. ## Pay for results. Outcome-based ### What is outcome-based pricing? Our pricing is tied to a business KPI: sales conversion, retention, autonomous resolution, whatever matters. Each task type has its own price, set together with the scope. [Let's talk pricing](https://cal.com/studiochat/25min) [Book a demo](https://cal.com/studiochat/25min) Your business KPI What you pay us They rise together: you pay on results. * Implementation included, no setup costs. * Integrations with the platforms already in use, included and with no hidden fees. * Hands-on support and continuous improvement of the assistant, included. HOW WE WORK ## From idea to a live agent. We don't just hand over software: we're there for the whole process. 1. 01 ### Process consulting We map out the process to automate together with the team, to define scope and expected impact. 2. 02 ### Integrations We connect the agent to current platforms (helpdesk, CRM, WhatsApp, whatever the team already uses), with no extra or hidden fees. 3. 03 ### Building and training the agent We help build the agent and train it on the team's playbooks, tone, and real cases. 4. 04 ### Ongoing enablement We train the team to keep optimizing the agent on its own, with or without SAMI's help. 5. 05 ### Follow-up We stay available to help whenever needed: new cases, process changes, whatever comes up. TEAM ## Team. [Ignacio Puig Moreno 01 Ignacio Puig Moreno Industrial Engineer (ITBA). Founder of Acamica (acquired). LinkedIn](https://www.linkedin.com/in/ipuigmoreno/) [Cristian Pereyra 02 Cristian Pereyra Computer Engineer (ITBA). Founder of Herolens (acquired) and ex-CTO at emblue. LinkedIn](https://www.linkedin.com/in/kshmir/) [Iván Itzcovich 03 Iván Itzcovich Computer Engineer (ITBA). Ex-Microsoft, ex-ASAPP and ML engineer at several AI startups. LinkedIn](https://www.linkedin.com/in/ivan-itzcovich-9aa7911b7/) FREQUENTLY ASKED QUESTIONS ## Questions, answered. ### What is StudioChat? StudioChat builds and runs AI agents for every kind of interaction with people: sales and support for customers, internal processes for employees, and coordination with suppliers and partners. The agent clones how a company's best talent already works (their tone, their playbook, and their judgment) and runs 24/7 across mobile SDK, in-app chat, WhatsApp, email, and voice. ### How is StudioChat different from a generic chatbot or an LLM wrapper? Most AI tools sell a generic chatbot and leave the team to build it alone. StudioChat captures how the best talent already talks and decides, using existing transcripts, knowledge bases, and playbooks, and delivers an agent that takes real action through the business's APIs, not just answers questions. We're also there for the whole process: it's not hiring StudioChat and being left alone to build and train the agent. ### What channels does the agent run on? Mobile SDK (native chat inside the app), in-app chat (web widget), messaging apps (WhatsApp Business), email, and voice. The same agent keeps one voice and knowledge base across every channel. In practice, it runs on any platform already in use, as long as it allows integrations. ### Does StudioChat replace Intercom, Zendesk, or Pylon? No. StudioChat connects to whichever helpdesk is already in use and works inside it, as one more non-human agent: it picks up cases, queries the business systems, takes actions through APIs, and hands off to a person with the full context when it should. What it doesn't replace is the omnichannel platform, and that's deliberate: the team keeps working where it already works, with nothing to migrate and no new tool to learn. ### How does handoff to a human work? It's triggered by an explicit request from the person in the conversation, negative sentiment, a skill outside its scope, or a VIP/critical account. The human gets the full conversation history, the data consulted, and the actions attempted, so they can pick up exactly where the agent left off. ### How is StudioChat priced? Outcome-based: it's priced per task resolved. The unit depends on the task, and each task type has its own price: resolving an inquiry doesn't cost the same as chasing a payment for three weeks. It can be a resolved conversation, a qualified lead, a closed sale, or a completed goal. Pricing is set together with the scope, tied to business KPIs (conversion, retention, autonomous resolution) instead of headcount. ### Can we start with one function and add others later? Yes, and that's how most accounts grow. You start with the task that hurts most, and once it works you add agents for other functions, on top of the same integrations and without switching tools. Coderhouse started by selling over WhatsApp and today runs three agents: pre-sales, student support, and staff support. You don't buy a package of N functions upfront: you add the next one once the previous one is already delivering. ### What metrics does StudioChat show? Built-in conversational metrics (sentiment, automation quality, recontact risk, CSAT/DSAT, autonomous resolution) plus whatever product KPIs get connected (sales conversion, churn, retention, NPS, activation), tied to every conversation. Reports can also be scheduled at whatever frequency is needed, delivered by email, WhatsApp, or whichever channel is preferred. ### How do I get started? Book a meeting with the team at https://cal.com/studiochat, or write to hey@studiochat.io. The first conversation maps the process to automate and sets the scope and the KPI the pricing is tied to. From there the team handles the integrations, builds the agent on your own playbooks, and stays on for training and follow-up. ## The playbook already exists. We multiply it. [Talk to us now](https://cal.com/studiochat/25min) [Book a demo](https://cal.com/studiochat/25min) [Or write to hey@studiochat.io](mailto:hey@studiochat.io)