The objections we get on every call, answered plainly. If yours isn't here, email us.
35|35|Lindy, Relevance, Stack AI are platforms. They give you tools, you build the agent yourself. You're hiring nobody — you're buying software. That's great if you have an in-house AI engineer.
45|45|Agentkin is the operator. We pick the platform, build the agent, run it, fix it, improve it. You're hiring a service, not learning a tool. Different products entirely.
46|46|It's the price of a part-time junior employee in NL/BE. For that, you get: an agent that works 24/7, never sleeps, never quits, never asks for a raise — plus a human operator (us) actively improving it every week. If it saves one person on your team 1 day per week, you've broken even on day one.
53|53|For context: a 3-month "AI transformation" project from Deloitte runs €100K+ with €20K/month managed services after. We're a small fraction of that.
54|54|Your agent runs on a dedicated VPS in a Netherlands datacenter — not shared with other customers, not on a SaaS multi-tenant system. Disk is encrypted, firewall locks everything down, daily security patches.
61|61|API credentials are stored in an isolated keystore that even the agent can't directly read (the agent uses them via internal channels). When you cancel, the entire VPS is wiped within 24 hours — we hand you a final backup of your data first.
62|62|We can sign your DPA, and we're GDPR-aligned by default since we're an EU-based operator.
63|63|No. On cancellation we hand you:
70|70|The agent runtime (Hermes) is open source. Anyone can host it. You're never locked in.
76|76|For any irreversible action (sending an email to a client, posting publicly, charging a card, deleting a file), we configure the agent to require human approval — either via the dashboard or a quick Telegram confirmation. You decide which actions are auto-approved vs. need-a-click.
83|83|For reversible reads (drafting an email, summarizing a doc, fetching data), the agent runs free.
84|84|Plus: every action it takes is logged in your dashboard. You can audit anything, anytime.
85|85|Today: Claude Sonnet 4.6 for most work, Opus 4.7 for hard reasoning, GPT 5.5 as fallback. We pay all the API costs.
92|92|In 6 months it'll be different. New models drop every quarter. We swap your agent to whatever's best — same day a better one ships. Your price doesn't change.
93|93|Our monitoring catches most failures before you notice — auto-restart on crashes, alerts to us within 60 seconds of anomalies. For things we can't auto-fix: typical response time on business hours is under 1 hour. Off-hours, under 4 hours for critical, next morning for everything else.
100|100|If you need 24/7 sub-15-min response, that's a separate engagement — talk to us.
101|101|Honest answer: it depends on the role and how you structure work.
108|108|What we see most often: agents replace the annoying parts of someone's job (inbox triage, follow-up reminders, report generation, data entry, basic research). The person now has time for the work they were actually hired for. Net positive for retention and morale.
109|109|If your goal is straight headcount reduction, we'd probably refer you to a different vendor. We build with humans, not against them.
110|110|Yes — we do live demos on every discovery call. We share our screen, show you the dashboard, show you a real agent running, and answer specific questions about how it'd apply to your work.
117|117|No pre-recorded sales theater.
118|118|Same price. The "unlimited agents" promise is real. In practice, most customers find that 2-3 well-built agents do the job of what they thought needed 10. Quality beats quantity.
125|125|Not at all. The industry list is where we have the most pattern recognition, so the first build will be fastest. But if you're a different shape (SaaS, e-commerce, consulting, finance, healthcare), book a call — we'll be honest about whether we're the right fit.
132|132|We don't take customers in regulated industries (healthcare, banking, etc.) right now because the compliance overhead doesn't fit our flat-fee model. We'll point you elsewhere.
133|133|Because the big agency will assign you a junior, charge you €25K/month, and use you to train them. Solo operators eat what we kill — your success is the entire business.
140|140|And honestly: this space moves too fast for a 50-person team to keep up. By the time their internal "AI Council" approves a new model, we've already deployed it for you.
141|141|Email us. We answer in plain words, usually within a few hours.
148|148| hello@agentkin.ai 149|149|