Front

Recently a customer’s COO was direct: “Instead of hiring another customer service rep, we are developing our own agent.” Then they skirted my question about where the budget was coming from: “I just want to get this thing working, then optimize for costs later.” 💡 This made sense. 0→1 agent development is cheap and easy, and COOs have slush funds for AI-related initiatives. But a COO also knows that 1→100 development is hard, eventually their agent is going to be expensive, and vendor lock-in is risky because their model provider could raise prices. So that became our wedge: Connect the agent you built. When you have clarity on costs, swap in our agent, or another provider’s. Your team stays on Front. Models will keep getting cheaper and better, so our goal is to enable our customers to swap agents as better agents arrive, without moving their team off Front. 👉 Learn all about our new external agent functionality my team built here: https://lnkd.in/ghCyFCc2

GLT Logistics processes 14,000+ emails a month 83% faster in Front. One inbox, clear ownership, and every customer gets a reply.

Electrical.com cut rep onboarding from 4-6 months to 2-4 and email response times from 4-6 hours to 1-2, with Front.

Today, Front took a bold stance on AI: "It's okay to see other agents" Most companies are pushing for "monogamy", "walled gardens", or [insert your preferred metaphor for a closed ecosystem here], even if their agent isn't the right fit for the job. And it makes sense why. Owning AI usage is more valuable for them, but it ends up leaving their customers in the lurch. Customer are forced to fill the gaps. They stand up custom agents in tools like Claude, Codex, or Gemini, which work but create a mess behind the scenes. Tons of systems. Tons of agents. No central way to govern and manage them. So we did two big things: launched custom agents in Front and opened up Front to external agents. Now, you can use our AI agent, Autopilot, or bring your own. And, you can manage, govern, and measure all your agents in one place. The result? One omnichannel workspace where teams and AI agents resolve customer requests faster, together. Read more about the launch and check out a cool demo video in the blog: https://lnkd.in/g88isAhh

71% of B2B companies reported significant AI issues in the past 3 months. That number gets attention. But the reason behind it matters more. Most of these failures aren't about the models. They're about the systems AI is running on. From our research: → 24% said AI created more coordination work → 22% lost context during handoffs → 20% saw routing failures These are coordination problems. AI can process a request. But in most environments, it can't see the full workflow around that request: Which teams are involved What's already been said What dependencies exist across systems So it makes decisions with partial context. And that creates downstream work for humans to fix. In other words: AI didn't introduce the coordination gap. It exposed it. If coordination lives outside the system, AI will always struggle. If coordination is built into the system, AI starts to work the way people expect. That's the real shift. And that's what Front was built for. Full research: research.front.com

Global's 7 operations teams now average a 1.5-hour response time. Automated routing gave managers ~20% of their time back.