Hatchium AI – WhatsApp support with AI, team inbox, and automation. · Fazier

Comments

32

We built Hatchium AI after watching businesses run sales and support from one employee's personal WhatsApp: missed messages, lost sales, and no team visibility. Hatchium gives that same business an AI assistant trained on its own content, a shared team inbox, and automations that recover carts, confirm COD orders, and rescue failed deliveries. Meta message rates pass through at cost with no markup. Happy to answer anything about the AI guardrails, the courier rescue flow, or the pricing model.

Passing Meta message rates through at cost with a live calculator is a smart trust move, how does the AI decide when to hand a conversation to the human inbox?

Thanks Zain! Three triggers: the customer asks for a human, the question can't be answered confidently from the business's own content, or it touches something the AI is not allowed to decide on its own (discounts, pricing promises, refunds, complaints). Those always route to the team inbox, and the agent gets a short AI-written briefing of the conversation so they aren't starting cold.

Hatchium AI Your product has strong potential, but I found a few key improvements that could make it even better. I'd love to share my feedback and suggestions—please contact me at [email protected]

Thanks Alex! Feel free to share the suggestions right here in the thread, happy to read them.

Cool app that helps people focus more on the real work than the mundane work that doesnt help me generate more income

Thanks! That's exactly the aim: the assistant takes the repetitive questions so your time goes to the work that actually earns.

Aditya Gudapati

Founder of ~ https://www.groupverse.co.in/ {groupverse} </> Free Group Links — WhatsApp & Telegram Communities

ok its good but not much imp i think sooo

Thanks for the honest take, Aditya. If you mostly sell or support over email it may not fit, but for teams whose customers live on WhatsApp it removes a lot of daily pain. Appreciate you looking!

The WhatsApp-first approach is smart given how many customers prefer chat over email. I am curious about the AI handoff - when the assistant escalates to a human agent, does it pass along the conversation context and sentiment analysis? That would save agents significant time. The transparent Meta message rate pricing with no markup is a nice differentiator too.

When the assistant escalates to a human agent the context is passed as summary to the Agent , and if it is a return/exchange matter then eligibility is also checked and mention in handoff of the customer is Eligible for refund/exchange

NA

Really impressed with Hatchium AI. The combination of WhatsApp AI support, team inbox, and automation makes customer communication much easier to manage. Simple and useful for growing teams.

Thank you Sagar! "Easier to manage" is exactly the bar we hold ourselves to for growing teams, glad it comes through.

Love how focused this is. WhatsApp support can get messy fast, so having the AI handle the repetitive questions while still making it easy to hand off to a real person seems like a great balance. Congrats on the launch!

Thanks Noah! That balance is the whole product: AI for the repetitive questions, a clean handoff with context when a real person is needed. Appreciate the kind words!

nice product very use full

Thank you Akash, glad it looks useful!

Hatchium AI looks really useful for businesses handling customer support on WhatsApp. Combining AI, a shared team inbox, and automation in one place could save teams a lot of manual work.

Thanks Anum! Cutting that manual back-and-forth is exactly what we built it for.

Rezycle.me Team

The at-cost Meta message rates is a smart trust move, most tools in this space hide their margin exactly there. One question about the cart recovery automations: can you cap how many times a single customer gets pinged? In my experience that's the point where these flows stop feeling helpful and start feeling like spam.

Thanks, and fully agree that's where these flows go wrong. Yes, it's capped: one recovery message per abandoned cart, plus a per-customer cooldown (24 hours by default, adjustable) so the same number is never pinged again even if the store creates multiple checkouts for that customer. Opt-outs are honored automatically on top of that, so a STOP ends it immediately.

Passing Meta's message rates through at cost instead of marking them up is the detail that'd get me to actually trust the pricing page. Most WhatsApp tools bury that markup and you only find it after you've integrated. The COD confirmation + failed delivery flow is a smart niche too, that's a real operational headache for ecommerce teams that most generic AI support tools don't touch.

Thanks Walker, this is exactly why we put the calculator on the pricing page: you see the actual Meta rate per conversation before you commit. And agreed on COD and failed deliveries, those flows came straight from watching ecommerce teams chase couriers by hand.

Hatchium AI looks like a practical tool for teams that rely on WhatsApp as their main customer channel. I like that it combines AI replies, a shared inbox, and automation instead of treating support as just another chatbot. The ecommerce use cases feel especially useful, from cart recovery to COD confirmation and delivery follow-ups. Transparent message pricing and a no-credit-card trial also make it easier to test. The strongest part is the human handoff, because customers still need real support when conversations get complex.

Thank you for such a thoughtful read of it! You're right that the handoff is where it matters most: the AI covers the routine, and when a conversation gets complex a real person takes over with the full context in front of them.

Founder of krooko.me

very good tool to be honest

Thanks Alex, appreciate you taking the time to check it out!

This looks practical for teams that already rely on WhatsApp for customer conversations. The strongest point is combining a shared team inbox with automation instead of making support jump between a bot and separate manual follow-up. I would add examples of escalation rules or handoff workflows.

Thanks, that's fair feedback on showing examples. Today it works like this: you set up named queues (Sales, Support, Billing) with their own staff, customers can pick the right department from a quick menu, and the AI escalates on its own when a question is outside your content or policy-sensitive, passing the agent a short briefing of the conversation. We'll look at putting worked examples of these flows on the site.

I like the WhatsApp-first approach, especially the combination of AI replies with a shared team inbox for human handoff. The cart recovery, COD confirmation, and appointment-booking workflows make it useful for both ecommerce and service businesses. I’d be interested to see how easily the AI can be trained and updated when business information changes.

The team inbox angle is smart. Most AI support tools pitch full automation, but small teams actually want the handoff: AI catches the message at 9pm, human takes over in the morning with context intact. One question from experience with local businesses: how does escalation work when the AI is unsure? A bad confident answer on WhatsApp costs more trust than a "let me get a human" ever does. If you let owners tune that threshold per topic, that would be a real differentiator.

Owners can tune the threshold, also they can Ban the bot from answering any money related topics so that bot does not end up making false promises.