Agent-Assist Tools for Customer Support Teams

𝗔𝗜 𝘁𝗵𝗮𝘁 𝗱𝗼𝗲𝘀𝗻'𝘁 𝗿𝗲𝗽𝗹𝗮𝗰𝗲 𝘆𝗼𝘂𝗿 𝘁𝗲𝗮𝗺. 𝗔𝗜 𝘁𝗵𝗮𝘁 𝗺𝗮𝗸𝗲𝘀 𝘁𝗵𝗲𝗺 𝘂𝗻𝘀𝘁𝗼𝗽𝗽𝗮𝗯𝗹𝗲. Here's what most companies get wrong about AI in customer support: They think it's all or nothing—either full automation or no AI at all. HubSpot just proved there's a better way. Their new Reply Recommendations feature is the perfect example of "AI assistance, not AI replacement." 𝗛𝗲𝗿𝗲'𝘀 𝘄𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀: → Support reps stay in control (review, edit, send) → Customers get faster, more accurate responses → Teams build confidence in AI gradually → No risk of rogue AI responses damaging your brand 𝗧𝗵𝗲 𝗴𝗲𝗻𝗶𝘂𝘀 𝗺𝗼𝘃𝗲? They merged Reply Recommendations into Customer Agent, making it a zero-risk entry point. Teams can experience AI's value without deploying a fully autonomous agent. 𝗪𝗵𝗮𝘁 𝗜 𝗹𝗼𝘃𝗲 𝗺𝗼𝘀𝘁:  • Reps can dismiss, edit, or use recommendations  • AI learns from your actual content sources  • No credits used (yes, really)  • Human expertise + AI speed = magic 𝗧𝗵𝗶𝘀 𝗶𝘀 𝘄𝗵𝗮𝘁 𝘁𝗵𝗲 𝗳𝘂𝘁𝘂𝗿𝗲 𝗼𝗳 𝘄𝗼𝗿𝗸 𝗹𝗼𝗼𝗸𝘀 𝗹𝗶𝗸𝗲:  • Not humans OR machines.  • Humans AND machines.  • The best teams won't be the ones who resist AI.  • They'll be the ones who learn to dance with it. P.S. If you're still debating "should we use AI?"—you're asking the wrong question. The right question is: "How do we use AI to make our people more effective?"

If chatbots talk, AI agents execute. What’s an AI agent? An AI agent is autonomous software that understands your goal, plans the steps, uses tools/APIs, and learns from feedback to finish the job with minimal supervision. Think proactive operator, not just a chatbot. 🧠🛠️ Why it’s a game-changer 🚀 - From replies to results: Books meetings, files tickets, reconciles data, triggers deployments, and verifies outcomes. - From tasks to outcomes: Orchestrates multi-step workflows and collaborates with other agents to hit KPIs. - From scripts to learning: Adapts to edge cases, remembers context, and improves every run. Real wins you can copy today ✅ - Customer Support: Auto‑triage tickets, search KBs, summarize history, propose fixes, and escalate only when needed. - Sales Ops: Prospect → qualify → personalize → schedule → update CRM without nudges. - Content Engine: Research → outline → draft → fact-check → repurpose for LinkedIn/IG/X → analyze and iterate. - IT/DevOps: Watch logs, detect anomalies, run playbooks, verify recovery, and post‑mortems—fewer 3 a.m. alerts. - Finance Ops: Reconcile transactions, flag anomalies, prep monthly close, draft stakeholder updates. How it works (simple loop) 🔁 Perceive → Reason → Act → Learn. Inputs in, plans made, tools called, results improved—on repeat. Start this week (no fluff) 🗂️ - Pick one repeatable workflow with clear success criteria. - List required tools/APIs (docs, CRM, ticketing, calendar, storage). - Set guardrails for autonomy vs. human approval. - Log everything; review weekly to tighten prompts, memory, and policies. Scroll-stopping openers 🎯 - “Chatbots answer. Agents deliver.” - “Outcomes > outputs. Meet AI agents.” - “One agent > five manual workflows.” 💬 Comment “AGENT” for a plug‑and‑play blueprint to automate your most annoying workflow this week. #AIAgents #AgenticAI #Automation #GenAI #LLM #ToolUse #Workflows #Productivity #CustomerSupport #SalesOps #DevOps #MLOps #AIinBusiness #Growth #Startups #APIs #Operations #Engineering #TechLeadershipa

Salesforce is all-in on AI Agents with Agentforce. Today at Dreamforce, Salesforce demonstrates Agentforce; a significant step toward leveraging AI agents to transform customer success, operational efficiency, and innovation across industries. Their tagline, "Humans with agents drive customer success," captures the idea of collaboration and synergies. Salesforce has a comprehensive approach to making AI Agents accessible and effective for all. ✅ A definition and framework for AI Agents. At its core, an Agentforce agent is defined by five key elements: - Role: The purpose of the agent on your team. - Knowledge: The data the agent needs to be successful. - Actions: The goals an agent can fulfill. - Guardrails: The guidelines an agent operates under. - Channel: The applications where the agent gets work done. ✅ Three agents available at launch: - Customer Service Agent: Enhances customer support with dynamic, conversational AI to handle inquiries 24/7. - Sales Development Agent: Engages prospects around the clock, acting as an intelligent sales agent that drives lead generation and customer acquisition. - Sales Coach Agent: Provides tailored coaching for sales representatives, including personalized feedback, pitch practice, and negotiation strategies, making every rep the best they can be. ✅ Additional agents coming soon: - Agentforce Assistant: Easily create and execute tasks for every employee. - Marketing Agents: Autonomously optimize and personalize marketing campaigns. - Commerce Agents: Instantly set up, manage, and optimize online storefronts. - Employee Service Agents: Automate onboarding and provisioning for new hires, enhancing internal operations. ✅ I see significant potential for Agentforce solutions in sectors like: - Banking: Automating credit risk assessments, fraud detection, and loan processing. - Retail: Managing inventory and order fulfillment, driving personalized marketing campaigns. - Healthcare: Supporting patient scheduling, providing medical assistants, and handling billing queries. - Manufacturing: Predicting equipment failures, optimizing supply chain management, and enhancing safety compliance. - Finance: Automating financial audits, reporting, and compliance monitoring. - IT: Modernizing legacy software systems, handling software requests, and performing security audits. 💡 While these first iterations of agents are powerful, they are still in the early stages, focusing on specific tasks without yet collaborating with other agents. However, this thoughtful and focused start lays the foundation for more sophisticated, interconnected AI systems in future versions—paving the way for a new era in how businesses leverage large language models and AI-driven solutions. I’m excited to see where Salesforce takes Agentforce next. Kudos to Salesforce for leading the charge in AI innovation and making these powerful tools even more accessible!

Email templates can help customer service reps improve efficiency. But what happens when just choosing the right one becomes overwhelming? It's a case where AI can unlock human super-skills. One company implemented an AI tool from Laivly to help agents select the right template. Laivly's "Smart Response" feature analyzes incoming emails to suggest the right template for agents to use. Agents can review the suggested template for accuracy, and add personalization before sending the final email. The Smart Response tool improved productivity by 49%. Even better, customer satisfaction increased 10% and first contact resolution rose by 17%. It's a great example of using AI to handle tedious, repetitive tasks so agents can be freed to concentrate on work where they can add more human value. I'm increasingly seeing stories like this. Rather than humans or AI, it's humans and AI.

Support teams face constant pressure to resolve cases faster without overloading engineering. For one Glean customer, valuable resources were tied up in avoidable tickets, MTTR (mean time to resolution) hovered at nearly two days, and agents spent hours manually triaging cases. Their goal: boost self-solves, improve MTTR, and reduce R&D reliance – without adding more tools. So they embedded Glean in Zendesk, giving agents prompts to quickly gather knowledge across all company data. In triage, agents use Glean to find similar tickets, summarize runbooks and past Jira investigations, and compile clear updates for customers or well-packaged escalations. That streamlined process now drives faster resolutions, smoother knowledge transfer, and consistent workflows—leading to: • 34% increase in self-solves with more future automation planned - this is incredible progress • 24% faster MTTR (1.9 → 1.5 days) • 2–4 hours saved per week for 85% of users (13–26 business days/year) • Reduced R&D involvement in lower-tier tickets By streamlining resolutions, knowledge transfer, and process consistency, the team achieved remarkable results – proof of what’s possible when AI is embedded into everyday workflows. Stories like this are energizing – showing how teams are using Glean to reimagine what they can accomplish.

Every week there's a new AI tool making the rounds. Claude Code. Open Claw. Hermes. Most business owners hear all the noise, don't know what's worth paying attention to, and end up doing nothing. Here's what's actually worth understanding: An LLM like ChatGPT or Claude is a car with no driver. It does nothing unless you get in and tell it where to go. Every prompt, every task, every decision runs through you. An agent is different. An agent has the driver already in the seat. It runs scheduled tasks, makes decisions, and completes work without someone prompting it every step of the way. You're not behind the wheel anymore. You're the advisor. The car runs without you. I run two businesses doing $650k/month combined, on track for $8M this year. Here are the 4 agents I'd recommend any service-based business build first. 1. A Sales Support Agent Lives in Slack, available to me and every member of the sales team. • It pulls transcripts from every call • Moves prospects through the pipeline • Adds CRM notes automatically • Handles follow-ups without anyone touching it. When someone closes, it fills out the new client form and triggers the enrollment sequence on its own. The closers' job is to be on calls. Everything else runs through the agent. 2. A Client Support Agent After every client call, it sends the recording, a summary, and action items directly to the client. If someone needs ad copy, the team tags it with the brief and gets 10 ready-to-run concepts in minutes. It also monitors client health, flags churn risk before it becomes a problem, and surfaces underperforming accounts before anyone has to manually check. 3. An Automated Onboarding Agent Kicks in the moment a client signs. • Analyzes their website • Identifies their branding • Builds a competitor analysis, creates ad creatives, and loads nurture sequences into the CRM — all before the kickoff call even happens. People decide within 48 hours whether they made the right buying decision. This agent compresses that timeline significantly. 4. An Internal and External AI Clone Internally, anyone on the team can ask it anything. It knows • Every SOP • Every process • Every training document we've ever built. Externally, clients ask it questions and it responds in my voice, trained on every call transcript and piece of content I've produced. When a client has a question at 11pm, they don't wait until the next call. They ask the agent. That's what digital leverage actually looks like. Don't start with the technology. Start with the friction. Write down everything that eats time in your business every week without requiring actual judgment. Find the most obvious one and build around that first. If you want our team to come in, audit every department, and build this out as your fractional AI department… DM me “AUDIT” We guarantee you'll save $50k in payroll or you don't pay.

At Chatbase, we’re only 15 people but we have 20+ AI agents running the show. Let me walk you through how we actually use them and why it works. 1. Social media is our first signal what’s trending, what’s changing, what’s clicking It’s not just about posting. We treat social like a listening tool. We’ve got agents that scrape LinkedIn, Meta, and Google ads from our competitors every day. They track what features they’re pushing, what pain points they highlight, and what language they use. That helps us spot early shifts in the market like when everyone suddenly starts talking about integrations or onboarding. So instead of guessing, I know what to post, write, or test next. We use: – Agent for scraping ads – Agent that summarizes tone, CTAs, strategy – Agent that drafts post hooks based on what’s working Tools: Lovable + Make..com + Supabase + OpenAI (We even have agent that helps me write my personal posts, more on that in a future post.) 2. Content is where the long-term game is played and we don’t play blind We mix two inputs: 1. What competitors are writing and ranking for 2. What our customers are asking and struggling with We have agents tracking competitor blogs and keywords so we know where we’re behind. But we also pull from our own changelog and support chats to turn real conversations into blog posts, carousels, newsletters, and help docs. We use: – Agent tracking competitors and keywords – Agent that surfaces support themes and insights from customer conversations (this is a really cool feature in Chatbase, you can summarize convos into groups. It’s great for support, but I love using it for marketing) Tools: Lovable + Make + Chatbase + OpenAI 3. Support isn’t a cost center, it’s our strongest growth engine Support is where we hear the truth. What’s broken? What’s unclear? What feature isn’t delivering? Our AI agents handle conversations directly on the site and inside the product but they also give us full visibility into what’s happening. And that’s gold for content, sales, and product. We use: – Agent that gives real-time support (with Stripe, Zendesk, Calendly, etc.) – Agent that turns FAQs into help docs – Agent that flags high-intent convos and routes to sales Tool: Chatbase 4. Sales isn’t just a funnel, it’s a feedback loop When someone reaches out for enterprise pricing, we don’t just want to “close” them. We want to know: – What pushed them to upgrade? – What blockers did they hit? – What feature made them reach out? This feeds back into our product messaging, pricing tiers, and content. We use: – Agent for qualifying leads – Agent for booking sales calls – Agent that logs which features are mentioned most during chats Tool: Chatbase That’s how we scale Chatbase. Not with a huge team. But with a smart setup. In the video, I wanted to show that every agent is connected and that delivering the best possible customer experience means marketing, sales, and support must work together.

The 3 types of AI tools every CS leader needs to understand (and how to use them) AI tools are everywhere, but as a CS leader, you need to cut through the noise and understand what actually matters for your operation. Here’s my simplified breakdown for customer success applications: 1/ Large Language Models (LLMs) What they are: The “brains” behind ChatGPT, Claude, Gemini - sophisticated tools that read and write like humans. How CS leaders use them: • Analyzing customer call transcripts to identify risk signals • Generating personalized QBR content based on usage data • Creating customer-specific success plans from templates • Summarizing months of customer interactions before renewal calls Key limitation: They don’t know your customer data unless you feed it to them. 2/ Workflow Automation Platforms What they are: Tools like Zapier, Workato, and Microsoft Power Automate that connect your existing systems and automate step-by-step processes. How CS leaders use them: • Automatically updating health scores when usage patterns change • Triggering alerts when customers miss onboarding milestones • Creating customer pulse reports by pulling data from multiple systems • Routing high-risk accounts to senior CSMs based on specific criteria CS-specific example: When a customer’s usage drops 30% week-over-week, automatically create a task for their CSM, pull recent support tickets, and generate a summary of their recent interactions. 3/ AI Agents *lWhat they are: Digital helpers that can complete specific tasks within larger processes, combining LLM intelligence with system integrations. How CS leaders use them: • Research agents that compile customer background before executive meetings • Health score agents that analyze multiple data sources to predict churn risk • Content agents that create personalized customer communications • Analysis agents that identify expansion opportunities based on usage patterns CS-specific example: An agent that monitors customer communications, identifies mentions of business challenges, researches relevant case studies, and drafts personalized recommendations for the CSM to review. —- I keep thinking about the ways to get started, it all seems like so much. Change management, getting IT or security involved… but you need to just start. Start with your biggest operational pain points: 1. Identify repetitive tasks your team does manually 2. Map which type of AI could address each task 3. Test with simple workflows before building complex agents 4. Measure impact in terms of CSM time saved and customer outcomes The technology exists today. The real work is understanding your CS processes well enough to determine where AI can replace tasks currently requiring human intervention. Remember: Agents handle individual smart tasks. Workflows organize how those tasks connect. LLMs provide the intelligence that makes it all possible. What CS process would benefit most from AI automation in your organization?

Stop calling it “just a chatbot.” That’s like calling an iPhone “just a flashlight.” If you’re a founder or operator in a small business, here’s how to actually use AI agents — mapped to real skill levels, real tools, and real outcomes. 🧭 Whether you’re no-code only, low-code curious, or already building multi-agent systems — this post is your practical roadmap. Let’s break it down: ⸻ 1️⃣ Beginner: No‑Code Agents for One Job ✅ Goal: Save 5–10 hours a week by automating obvious, repeatable tasks. 🛠️ Common use cases: • Website support agents (Chatbase, Intercom, Zendesk bots) • FAQ or “concierge” chat for pricing and booking • Internal helpers for SOPs, doc search, or meeting summaries 🎓 Key skills: • Prompt & role design (clear tasks, handoff rules) • Connect tools with no-code builders (Voiceflow, Make, Chatbase) • Plug into knowledge bases (Notion, PDFs, Help Centers) 🏁 Output: 1–3 agents that answer questions, collect info, and hand off to humans. ⸻ 2️⃣ Intermediate: Agents That Do Work ✅ Goal: Go beyond Q&A — automate workflows that move deals, tickets, or cash. 🛠️ Use cases: • Lead qualification and routing (RevOps) • Follow-up automation and CRM actions • Invoice generation, payment reminders • Tier 1 support and ticket resolution 🧰 Tool stack: Make, Zapier, n8n, CRMs, Helpdesks, Stripe, Google Sheets 🎓 Skills to level up: • API wiring and workflow logic • Light RAG (retrieve and use custom content) • Guardrails: what the agent can and cannot do • Logging and simple A/B tests for prompts 🏁 Output: Measurable time saved on lead handling, customer support, and ops. ⸻ 3️⃣ Advanced: Multi-Agent Systems That Run Like Teams ✅ Goal: Deploy “always-on teammates” that act like a research analyst, content creator, or CS pod. 🛠️ Use cases: • A full RevOps copilot that researches, writes, analyzes pipeline • Multi-agent support pods with decision trees and policy enforcement • Forecasting agents for pricing, planning, and scenario testing 🧰 Frameworks: LangGraph, CrewAI, AutoGen, LangChain, Semantic Kernel 🎓 Advanced skills: • Agent frameworks and orchestration • Custom internal tools and APIs • Real governance: permissions, approvals, observability 🏁 Output: Persistent agents that think, plan, and act within policy — just like a real employee. ⸻ 🔁 Each level builds on the last. You don’t have to leap into LangGraph tomorrow. But if you’re still stuck in the “FAQ chatbot” phase — now’s your cue to upgrade. 💬 Drop your niche or stack below and I’ll suggest a 90-day agent roadmap that matches your tools and goals. ➕ Follow for more system-level AI strategies you can actually use.

We built a Zendesk email assist AI agent and it's handling a full quarter’s work for one human support rep. Here's the step-by-step flow: 1. User sends a complex or nuanced product question to support@voiceflow.com 2. Tico (our AI agent) reviews the question and passes the content and intent. 3. The most fitting knowledge base is tapped via confidence level. 4. A personalized, accurate & highly-specific response is drafted. 5. The draft is slotted into Zendesk as a private comment. 6. Our team reviews, tweaks if necessary, and sends it to the user. This has slashed the onboarding and training time for support staff that's typically slowed down by the complexity of the product. The impact? ✅ Our support team is no longer just keeping up; they’re ahead, delivering faster, sharper responses. ✅ Customers feel understood, their issues addressed with pinpoint accuracy, boosting our CSAT scores. ✅ Tico’s continuous learning means every interaction makes it smarter, ready for even the most nuanced queries. So far, Tico Assist is tackling over 2000 tickets - a full quarter’s work for one human support rep, for less than the price of lunch. If you’re navigating high support volumes with a lean team, this type of Zendesk AI Assist Agent can help blend automation with quality for your customers. P.S. Tico doesn’t just fetch any answer. It pulls from the most relevant knowledge base (e.g. a technical code response for a developer question). From my post last week, this multi-knowledge base strategy is something that I think we will see much more of in CX this year.