What 20 months of building looks like :)
Hey PH 👋 Kshitij here from Inrō.
We launched here back in 2024, which gave us a real platform to kickstart our journey, and the feedback from this community has genuinely shaped how we built Inrō.
A lot has shipped since then, and we're back on Saturday, April 25, with our biggest update yet.
But before that, here's everything we shipped in these last few months -
Automation triggers: From 6 to 20+ trigger types. Posts, Reels, Stories, IG Live, mentions, inbound DMs, referral links, Meta ads, external webhooks, and more.
Automation actions: Grew from ~10 to 40+. You can now copy-paste blocks between scenarios, duplicate them, bulk-delete them, and share flows as templates. Plus, a simplified builder that makes getting started much faster.
Scheduling and retroactive automations: All your IG content in one place. Run automations on specific posts, schedule them for upcoming ones, or trigger them retroactively on comments up to 7 days old.
All of these combined mean that whether you run a simple comment-to-DM flow or a multi-step funnel for all of Instagram, we have you covered.
Campaigns: Multimedia sends, audience customization, scheduling, branching logic, and a full template library across industries so you're never starting from scratch.CRM:
Custom properties, 30+ filters to segment your audience, and insights like never before. All captured automatically, fully exportable, and connectable to your existing tools.
Safety and quality controls: Spam detection, hate comment filtering, viral comment moderation that adjusts as your post blows up, opt-out detection, manual interjection, folder exclusions, and time-of-day scheduling. The stuff that quietly saves you from yourself.
Integrations: Native connections with Shopify, Calendly, Stripe, and ElevenLabs. 8,000+ tools via Make and Zapier. Full private API, two-way webhooks, and an MCP server so you can run Inrō from Claude or ChatGPT in plain language.
On the traction side: we started with a 10 beta users and we're now at 10,000+ creators, brands, and e-commerce stores using Inrō across all kinds of industries and use cases. Walmart, Kering, Publicis Groupe, and Virgin Voyages came on board along the way.
Whatever you're trying to build, someone in our user base has probably done it already.
Now, this is just part 1. The real shift is what we've built on top of all of this. Inrō is becoming an AI-first platform, meaning you get everything above without setting most of it up manually and without paying anything extra, unlike other tools.
The full AI launch is coming Saturday April 25 with surprises, bonuses, and a lot of new features. I'll be sharing more details in the week leading up to it, so stay tuned :)
One question before I go: What part of Instagram automation do you feel should already handle itself, but still doesn't?
Drop it in the comments. If we don't have it yet, it might just become our next feature ;)
Here was the launch -
https://www.producthunt.com/products/inro/launches/inro-ai
Replies
Are most users using simple flows or complex funnels?
Inrō
@freya_madison great question!
Actually, it breaks down into three pretty distinct groups.
The majority are running simple, high-volume flows. Comment-to-DM, story reply automation, that kind of thing. Set it up once, let it run. For these users, we built a simplified flow builder to make it quick to get started with these quick flows.
The second group combines multiple actions and multi-step flows, and this is where templates have become a big deal. Many creators and users we partner with will build a solid scenario and share it as a template, giving others a much stronger starting point than a blank canvas.
Then there's a smaller but fascinating third group doing things we didn't fully anticipate. Full end-to-end Instagram funnels, multi-scenario setups, external integrations, and some even controlling Inrō via MCP through Claude or ChatGPT.
What's interesting is that people move between these groups faster than you'd expect. Someone running a one-step flow in week one is usually on multi-step funnels with CRM capture and follow-ups by month two. And complexity tends to follow use case, not user type. A solo creator launching a digital product often builds something more advanced than a mid-sized brand just answering FAQs. It's less about sophistication and more about what they need to get done.
All three groups are using the same platform. It just grows with you.
What category do you think you fall into?
This is one of the most honest “real building” posts I’ve seen here.
The jump from features → systems is where most products actually change.
Inrō
@judit10 Thanks. Curious on what caught your eye the most?
Are users customizing templates heavily or using as is?
Inrō
@yara_simone Mix of both, but there's a clear pattern.
Most users start with a template and customize from there. The template handles the structure, and they swap out messages, add conditions, connect their CRM properties, and adjust timing. A coach tweaking a lead gen flow or an e-commerce brand adjusting a product drop sequence. Close enough to be useful, flexible enough to make it theirs.
Building from scratch tends to be for more advanced users who know exactly what they want, or for people solving something very specific. For example, setting up a single action to email themselves whenever a new Instagram lead comes in. Small, precise, not worth finding a template for.
The most interesting behavior we see is when someone builds something from scratch, gets it working well, and shares it as a template for others. That's honestly become one of the better ways good flows spread through the community.
That’s a huge amount of shipping in 20 months. The part that stands out is how the platform is moving from individual automations to more complete workflows across the funnel.
One gap I still see in Instagram automation is handling context across interactions. For example, connecting a user’s comment, DM, and past engagement into a more continuous flow instead of treating them as separate triggers.
Curious how you’re thinking about cross-channel or multi-touch context as you move toward the AI-first layer.
Inrō
@jahnavi_thota
Really sharp observation and honestly one of the more interesting problems in this space.
Every contact in Inrō already has a full history, including their comments, DM interactions, folder memberships, and custom properties, all tied to one profile. So when someone DMs you after commenting on three posts, the AI Agent isn't starting from scratch. It already knows who they are and how they've engaged.
What we don't do perfectly yet is stitch those touchpoints into a single, continuous flow automatically. Right now, you can build toward it by using contact history as conditions, routing based on past behavior, and updating CRM properties at each touchpoint. But it still requires you to design it intentionally.
The AI-first direction we're announcing is directly aimed at this gap. The vision is that Inrō holds the full context of every interaction across every surface, including comments, DMs, story replies, and campaign responses, and that the AI uses that context to respond and act in a way that feels continuous rather than each trigger being an isolated moment.
It's one of the harder problems to solve well because the data is there; the challenge is getting the AI to use it in a way that feels natural rather than creepy. Getting that balance right is something we've been thinking about a lot.
If you had this working as described, what would you build? Would love to hear your specific use case to guide us.
Congrats on 20 months, Kshitij. Going from 6 to 20+ trigger types is a lot of surface area to manage while keeping the product coherent. Respect.
To your question, the part that should handle itself but still does not is knowing when NOT to send. Most IG automation tools optimize for volume: more DMs, more replies, more reach. But the best responses often come from staying quiet on certain posts and being selective about when to engage.
A tool that could learn "this comment thread is going sideways, skip it" or "this user's tone suggests they are already annoyed, do not auto-reply" would be something I'd pay for. Sentiment-aware suppression, not just sentiment-aware targeting.
Building something adjacent to this at ad-vertly for the broader marketing automation space, so I'm watching what Inro ships closely. Good luck on the April 25 launch.
Inrō
@gaurav_singh91
This is such a sharp way to put it, and I suspect you know this pain firsthand from building in the broader automation space.
We've actually built a lot around precisely this.
Folder exclusions let you carve out anyone you don't want touched by automations entirely
AI-detected conditions mid-flow can read tone and intent before deciding whether to continue
Stop when manually entered pauses everything the moment a human steps in
Run once / run once in X time prevents the same person from getting hammered by the same flow repeatedly
Scenario priority groups mean if multiple triggers fire at once, only the one you've ranked highest actually runs
Automatic opt-out detection catches disengagement signals globally without any setup
The "this thread is going sideways, skip it" use case is something our hate and spam filter touches.
But sentiment-aware suppression as a standalone concept, knowing when silence is the right move, is honestly a great way to frame where this goes next.
Watching what you're building at ad-vertly too. Would love to compare notes sometime :)