n8n AI Support Automation -- 2

$25-50 USD / hour

Posted 1 minute ago

•

Ends in 6 days

$25-50 USD / hour

I need a complete n8n workflow that lets two large-language models—Claude and Google Gemini—jointly run my customer support. The agents must monitor both Email and my embedded Website chat, draft helpful replies in real time, and send them automatically whenever confidence is high. Key flow • Incoming message arrives via IMAP or the chat widget’s webhook. • The text is routed through a prompt chain that lets Claude propose an answer and Gemini cross-check it. • If both models agree above a confidence threshold, the reply is sent back through the same channel. • Whenever either model flags low confidence, the full conversation (including suggested answer and metadata) is pushed to a human teammate through a dedicated Slack channel I will provide. What I expect to receive • A ready-to-import n8n workflow file. • API key fields clearly separated so I can drop in my own credentials for Anthropic and Google AI Studio. • Clear, commented nodes showing where prompt templates live so I can tweak tone or policy wording later. • Brief README or loom video walking me through setup, environment variables, and how to adjust confidence thresholds. Acceptance test If I create a sample email asking an unusual product question and the models decline with a “Need human” flag that lands in Slack, the job is done. Likewise, a common FAQ sent by chat should return a correct, well-formatted answer within ten seconds. Please keep the solution entirely inside n8n (no external servers) and stick to native or HTTP Request nodes only—no custom TypeScript modules this time.

Project ID: 40823122

About the project

65 proposals

Open for bidding

Remote project

Active 5 secs ago

Place your bid

Bid amount Email address

Benefits of bidding on Freelancer

Set your budget and timeframe

It's free to sign up and bid on jobs

65 freelancers are bidding on average $34 USD/hour for this job

User Avatar
Flag of UNITED STATES

Hi, I am Elias from the sunny state of Florida. I understand you're looking to create a complete n8n workflow that integrates Claude and Google Gemini. The main goal here is to leverage their capabilities efficiently while ensuring smooth communication between them. To approach this, I would start by outlining the data flow and how these AI models will interact within the n8n framework. We can utilize n8n’s HTTP Request nodes to connect to both APIs, ensuring proper authentication and error handling. I’d also set up triggers that allow one model's output to feed directly into the other, optimizing the workflow for responsiveness and efficiency. A few questions to clarify the scope: Q1 – What specific tasks or queries do you want the models to handle collaboratively? Q2 – Are there any specific conditions or triggers you envision for initiating the workflow? Q3 – How do you plan to handle the output from both models? Looking forward to hearing from you.

User Avatar
Flag of NIGERIA

Hi, Are you handling inbound email via IMAP idle polling or a webhook relay, and should the dual-model arbitration node evaluate raw token logprobs or a structured JSON confidence score? I will structure the n8n workflow to ingest IMAP and webhook payloads, pass the context through Anthropic and Google AI Studio nodes for dual evaluation, and parse the resulting confidence thresholds cleanly. Building multi-agent decision pipelines with failover routing to Slack is work I handle daily, so I will configure the prompt chain to execute sequentially, validate agreement, and trigger either automated dispatch or human handoff seamlessly. For $25, delivery in 1 day includes the complete, importable n8n workflow file with parameterized environment variables, commented prompt nodes, and a setup README. Tell me your preferred IMAP polling interval so I can dial in the trigger node right away.

User Avatar
Flag of PAKISTAN

Hello, Building AI-powered automation platforms and conversational AI solutions is a core part of Live Experts’ work, backed by over a decade of technical delivery experience and hundreds of completed projects. I can create your n8n-based support workflow connecting Claude and Google Gemini for email and website chat, with automated replies for high-confidence cases and Slack escalation whenever either model returns a “Need human” decision. My experience in AI/ML, AI chatbots, workflow automation, cloud APIs, and system integrations aligns directly with your requirements. I will keep the implementation inside n8n using native and HTTP Request nodes, with clearly separated API credential fields, commented prompt sections, conversation metadata, confidence handling, and channel-specific responses. My software architecture, JavaScript, API, and HTML skills also support clean, maintainable integration with your existing support channels. I will include the ready-to-import workflow and concise setup documentation covering environment variables, credentials, prompt updates, and confidence thresholds. Live Experts is ready to deliver a practical, production-focused automation that meets your email, chat, Slack, and acceptance-test requirements. Thanks!

User Avatar
Flag of INDIA

Hi, I can build this entirely inside n8n using IMAP/Webhook triggers, Claude for draft replies, Gemini for independent verification, and an IF-based confidence gate to auto-send only when both agree. Low-confidence cases will package the full thread, metadata, and suggested answer into Slack for human review. I’ll deliver an importable workflow, clearly separated credential fields, and commented prompt nodes for easy tuning. Please start the chat so we can discuss your channels, thresholds, and Slack handoff details further. Regards, Sai Bhaskar

User Avatar
Flag of PAKISTAN

Hi there, You need a robust n8n workflow that allows Claude and Google Gemini to handle customer support efficiently while ensuring replies are drafted in real-time. My experience with AI integrations and n8n makes this a fitting project for me. How I would approach it: - Create a flow that connects receiving messages via IMAP or your chat widget’s webhook to the respective language models. - Design a prompt chain for Claude and Gemini to evaluate responses, implementing a confidence threshold for auto-replies. - Set up fallbacks to push low-confidence cases to Slack for human review. - Deliver an organized workflow file with clearly defined API fields, commented nodes for easy customization, and a brief guide for setup. How will both models handle nuanced queries that might require additional context? I’m ready to deliver a solution you can rely on for seamless customer interaction.

User Avatar
Flag of TURKEY

Hi there, I see you're looking to develop a complete n8n workflow to connect Claude and Google Gemini for your customer support automation. I can create a robust system that captures incoming messages via IMAP or the chat widget’s webhook and efficiently routes them through a prompt chain for real-time drafting of replies. Ensuring both models agree before sending responses will be prioritized, alongside effective management of low-confidence flags being sent to your designated Slack channel. I will provide a ready-to-import n8n workflow file with separate API key fields for easy integration of your credentials. Additionally, I will include clearly commented nodes to facilitate adjustments to prompts, along with a brief README or instructional video covering setup and confidence threshold adjustments. I'm confident that my skills in API integration and chatbot development align perfectly with your requirements. Regards, Ali

User Avatar
Flag of INDONESIA

Hi there, I’m excited to propose a robust n8n workflow for your dual LLM AI support automation. Your vision for Claude and Google Gemini working collaboratively to handle both email and website chat is both forward-thinking and fully achievable within n8n’s native capabilities. I understand the workflow needs: IMAP and chat webhook integration, prompt chaining where Claude drafts and Gemini cross-checks, confidence threshold logic, and seamless escalation to Slack for human review when needed. I’ll ensure API keys for Anthropic and Google AI Studio are clearly separated, so you can easily manage credentials. Each node will be thoroughly commented, especially around prompt templates, making it simple for you to adjust tone or policy language in the future. My experience includes building advanced n8n automations leveraging LLMs, integrating various APIs, and designing reliable multi-agent support systems. I specialize in creating maintainable, no-code workflows using only n8n’s native and HTTP Request nodes, as you requested. Here’s my approach: I’ll develop a modular, ready-to-import n8n workflow with clear variable placeholders and logic for confidence scoring. The Slack escalation pathway will include all relevant metadata and conversation history, and I’ll provide a concise README and/or a Loom walkthrough covering setup, variables, and threshold adjustments. I’m committed to delivering a solution that’s easy to maintain, highly reliable, and meets your acceptance criteria. Looking forward to collaborating to streamline your customer support process! Best Regards

User Avatar
Flag of UNITED KINGDOM

Hello, Building a dual-model validation loop entirely inside n8n requires careful management of state and sub-workflows to prevent execution bottlenecks between the IMAP poller and the chat webhook. I will structure the workflow using native n8n nodes, combining an upstream trigger router with an HTTP Request node chain that queries Claude first, passes the draft to Gemini for cross-checking, and evaluates the resulting confidence JSON against your threshold. When consensus fails, the execution path will branch cleanly to format the conversation context and push the alert payload to Slack using native webhook nodes. Are you planning to handle chat session history using n8n memory nodes or persist the conversation state directly in your database? Have a nice day

User Avatar
Flag of UNITED STATES

I can build this directly in n8n using native nodes and HTTP Requests, integrating Claude and Gemini for answer generation and cross-validation. I’ll handle email and website chat inputs, confidence-based auto-replies, and Slack escalation whenever human review is needed. You’ll get a ready-to-import workflow, clearly organized API credentials, editable prompts, and a short setup guide. I’ll also test both the FAQ and low-confidence scenarios before delivery.

User Avatar
Flag of PAKISTAN

I will build a complete n8n customer-support workflow connecting Email and your website chat to Claude and Google Gemini for response generation, cross-checking, confidence scoring, and automatic replies. I’ll implement human escalation to Slack whenever confidence is low, keep everything within native n8n and HTTP Request nodes, and provide clearly organized credential fields, editable prompt templates, configurable confidence thresholds, and proper error handling. You’ll receive the ready-to-import n8n JSON plus concise setup documentation or a Loom walkthrough, and I’ll test both the automated FAQ response and low-confidence Slack escalation against your acceptance criteria. Best Regards, Muhammad

User Avatar
Flag of INDIA

I can build your dual-model customer support workflow entirely inside n8n using native and HTTP Request nodes, ensuring seamless integration between Claude and Google Gemini without any external servers or custom TypeScript modules. My approach involves setting up the IMAP and webhook trigger nodes to capture incoming messages from both your email and website chat widget. I will configure the prompt chain nodes so Claude proposes the initial reply and Gemini cross-checks it against your confidence threshold. If they agree, the response will automatically route back through the correct channel. If either flags low confidence, the workflow will format the conversation history, suggested answer, and metadata, and push it directly to your designated Slack channel. I will clearly separate your Anthropic and Google AI Studio API key fields and include commented prompt templates along with a brief README walking you through setup and threshold adjustments. Let's discuss how we can get this running quickly.

User Avatar
Flag of PHILIPPINES

Hello, I will create a complete n8n workflow that integrates Claude and Google Gemini for your customer support, enabling real-time monitoring and automated replies based on confidence levels. Using n8n's native nodes and HTTP requests, I recommend structuring the workflow to efficiently route messages, validate responses from both AI models, and manage low-confidence scenarios for human intervention. I will deliver a ready-to-import n8n workflow file with clear API key fields and commented nodes for easy customization. Testing will confirm functionality with example inquiries, ensuring responses are timely and accurate. Please confirm your preferred confidence thresholds and any specific prompt templates you want to use. Please feel free to reach out to me. I look forward to working with you. Best regards, Billy Bryan

User Avatar
Flag of AUSTRALIA

Hello, I can build this n8n automation so Claude and Gemini work together exactly the way you described: reading incoming email and website chat, drafting responses, cross-checking each other, and either replying automatically or escalating to Slack when confidence is low. I’ve recently built a similar n8n support flow where one AI drafted the response and a second model verified it before sending, with a clean fallback to human review for edge cases. I can structure this in a way that stays fully inside n8n, using only native/HTTP Request nodes, with clearly separated API credential fields and well-commented prompt sections so you can easily edit tone, policy, or thresholds later. I’ll also make sure the workflow is practical for your team to use, with a simple README or Loom walkthrough covering setup, environment variables, and how the confidence logic works. A couple of quick questions: 1. Do you already have the exact Slack channel webhook/target ready for escalation? 2. Should the website chat replies preserve the same formatting as email, or use a shorter chat-friendly style? If helpful, I can also suggest the best confidence-check structure so the handoff stays reliable and fast. Happy to jump on a call to align the flow before I start. Best regards, Mehreen

User Avatar
Flag of PAKISTAN

I got you! I can build your complete n8n customer-support workflow where Claude drafts, Gemini verifies, high-confidence replies auto-send via Email/chat, and low-confidence cases escalate to Slack with full context. I’m ready to handle the exact workflow structure, commented prompt nodes, separated credential fields, README/Loom guidance, and threshold tuning. I’m young, a fast learner, available 24/7, and focused on delivering this cleanly inside n8n using only native and HTTP Request nodes. Get the demo first before you pay. A couple of quick questions: Which chat widget/webhook format are you using, and should replies preserve full email threading headers for IMAP/SMTP? Also, do you already have an FAQ/knowledge base source to include in the prompt, or should I design placeholders for it? Let’s chat and discuss these answers so I can start right away. Kind regards, Haroon Z

User Avatar
Flag of MALAYSIA

Two models arguing over who answers the ticket is the fun part. I would run Claude and Gemini through separate n8n branches, then a judge node that merges or escalates when they disagree, so the client gets one clean answer instead of two conflicting ones. I can start today and have a working flow by Friday. Budget and timeline here are rough, based on the post as written, we will firm them up once I see your actual support channels and data sources. Want me to send a quick scope doc?

User Avatar
Flag of UNITED STATES

Hi there, Your n8n workflow needs to coordinate Claude and Google Gemini across email and website chat, automatically answer only when both models meet the confidence requirement, and escalate uncertain conversations to Slack without external servers or custom TypeScript. I have strong hands-on experience with Software Architecture, Java, and Automation and have solved similar problems involving AI workflow orchestration, webhook routing, and human-in-the-loop support processes. My approach would be to: * Review the required IMAP, chat webhook, model, and Slack integration behavior. * Identify the root cause of possible routing, confidence, and channel-reply issues before implementation. * Implement a clean, maintainable solution using native n8n nodes and HTTP Request nodes, with clearly separated credential fields. * Test edge cases, reliability, performance, security, and integration behavior, including the ten-second FAQ response target. * Deploy or integrate the fix without disrupting the existing system, including a ready-to-import workflow and setup documentation. I will structure the prompt chain so Claude proposes, Gemini cross-checks, and deterministic IF logic handles agreement, confidence thresholds, metadata preservation, and human escalation. Keeping credentials and editable prompts isolated also makes future policy changes safer. I prefer fixing the actual root cause rather than applying temporary patches, and I will keep the implementation simple, stable

User Avatar
Flag of INDIA

⚠️ If you're not happy, you don’t pay. ⚠️ Hi Client, thank you for checking my proposal and sharing the detailed project brief. I can build your n8n workflow using native nodes and HTTP requests with a scalable and efficient design. I will deliver: • IMAP integration to capture incoming messages seamlessly. • Webhook setup for real-time chat monitoring. • Prompt chain for Claude and Gemini to collaborate on responses. • Confidence threshold logic to automate replies or escalate to Slack. • Customizable prompt templates with clear annotations for easy adjustments. • A comprehensive README and Loom video for setup guidance. You will also receive: • Ongoing support during implementation to ensure smooth operation. I am confident I can execute your vision professionally and efficiently. Looking forward to discussing the timeline and next steps. Best regards, Chirag Pipal

User Avatar
Flag of VIETNAM

Hi, I will build your complete n8n support automation using Claude and Gemini, with no external servers or custom code. I will integrate email and website chat, configure both AI models to review responses, and automatically escalate uncertain cases to Slack for human support. I will also provide a clean, reusable workflow with editable prompts, secure API credentials, and clear setup documentation. I'd be happy to bring this automation to life! Best regards, Manh

User Avatar
Flag of INDONESIA

claude and gemini checking each other before anything goes out to your customers is a setup i can build for you in n8n i can take this on for you end to end, with the slack handoff to your teammate included i'll stay on it until it runs the way you want it to, and fix anything that breaks along the way which chat widget is on your website right now? happy to chat more whenever suits you Sela

User Avatar
Flag of INDIA

Building this dual-agent n8n workflow requires a clean router node to handle the IMAP and webhook triggers, followed by parallel HTTP request nodes for Claude and Gemini. I will implement a logic gate using a code node to compare confidence scores before triggering the Slack alert or the reply. Are you using a specific JSON structure for the chat widget webhooks, or should I build the workflow to normalize generic payloads?

内容来自 Freelancer News,版权归原作者所有。

阅读原文 ↗

把喜欢的内容,留在你的新标签页。

打开我的订阅