This is interesting, @dudabardavid1👀 I’ve tried Superhuman and a couple of other AI-driven email assistants, but most of them either lock you into one AI model or feel disconnected from Gmail’s native flow.
How do you see HeyHelp stacking up against tools like Superhuman or Shortwave, especially for founders who prioritize speed of triage over deep personalization?
Also curious: which use case have early users said saves them the most time so far, auto-sorting or reply drafting?
@rivi_richards thanks for raising this! For founders who care most about speed, auto-sorting/triage usually makes the biggest impact. Inbox noise drops instantly. When comparing to Superhuman or Shortwave, the two key differences you've already mentioned:
Native in Gmail, which brings a bunch of advantages like 2-way draft and label syncing, compatibility with existing extensions, and avoids some of the workflow issues users often mention with those tools.
No lock-ins (or markup pricing!)
From what we’ve seen so far, most users would agree that auto-sorting is currently the top time-saver.
@dudabardavid1 @rivi_richards And we also are the only company that allows you to choose your own API provider, so if you prefer to use Anthropic because of their human like emails, or chatgpt because you are used to that, we allow you to, those other providers force you to choose their choice of AI, liking you or not, that's the only option, we give you the flexibility to choose your preferred one.
The BYOK approach is smart - lets people control their AI costs and switch models based on what works best for their use case.
How accurate is the cold pitch detection? We get tons of outreach emails and most filters either miss obvious spam or flag legitimate partnership inquiries. The balance is tricky.
Also curious about the voice learning - does it pick up on writing patterns from sent emails, or do you need to train it manually on your style?
@alex_chu821 glad you asked. The pitch detection/auto-archive is actually one of my favorite features (I get tons of pitches every day too). HeyHelp looks at signals in the email body (e.g. wording like “discovery call”) and the sender (e.g. someone you’ve never interacted with before). Occasionally a pitch still lands in my inbox, but it’s much rarer than a legitimate email being archived by mistake.
As for learning, it automatically adapts from your sent emails, no manual training required. You can also fine-tune it in the AI Knowledge section by adding FAQs, style notes, or do’s and don’ts (e.g. “no meetings on Fridays”) to cover things your sent emails alone wouldn’t teach it.
@dudabardavid1 The signal-based detection approach sounds solid - using sender history plus content patterns makes sense. Better to have the occasional pitch slip through than lose real opportunities.
The automatic learning from sent emails is clever. Most tools require you to manually train them which nobody has time for. The AI Knowledge section for edge cases like scheduling preferences is a nice touch - covers those personal quirks that don't show up in email patterns.
Does it handle different types of legitimate cold outreach well? Like if someone's reaching out about a genuine partnership vs pure sales pitch?
@alex_chu821 Not at the moment, right now cold outreach is all treated the same, so partnership offers may get filtered alongside sales pitches. That would be the next level!
@dudabardavid1 That's the tricky balance with any automated filtering - legitimate partnership opportunities often use similar language patterns as sales pitches. Both tend to mention 'collaboration' and 'mutual benefit' type phrases.
Maybe a whitelist feature for domains or senders you've interacted with before could help distinguish between random cold sales and actual business inquiries. The context of previous interaction matters a lot for determining intent.
@alex_chu821 I'm using it myself as well, and not only HeyHelp categorizes/tags the pitches well, but it also auto archives them, in the last month probably just missed a couple emails that should be classified as pitches, but it didn't, all the rest (over 100) were correctly filtered from my inbox directly to the label.
Also about the knowledge base, if you add information there you give it some directions, but after almost 3 weeks using the smart reply, I'm finding it better and better everyday, because it's learning from every change that I make on the drafts and also with the few emails that I write manually myself, so it's getting smarter and closer to understanding what I need and want for all email replies! The self learning is really working great.
Incredible work! You've struck a really nice balance of power and simplicity.
I was wondering whether in addition to organizational groupings, whether HeyHelp could figure out sentiment or urgency too? Or perhaps some way to determine the complexity of the email — whether it's a quick 30 second job, or if it'll take me all afternoon to reply.
@buzzusborne thanks for the feedback. Not at the moment, but it's a good idea! Funny enough, we're about to release sentiment detection on our other product DragApp (which is mostly used by customer-facing teams, hence the focus on monitoring sentiment), but could be a good fit to HeyHelp too. You can follow our roadmap or share more ideas here.
Email overload is such a universal pain point, and this feels like a really practical solution. Congrats on your launch!
@christyfea thanks Christy, appreciate that. Inbox overload eats up so much time... We wanted a solution that actually gives people hours (and headspace) back each week. Glad it resonates!
@christyfea that's the plan! AI follow ups and AI calendar scheduling (done for you) is next!
@christyfea We have more to add in the next releases: AI follow ups, Custom Tags, and AI calendar scheduling, and so many others!
HeyHelp turns Gmail into your smartest assistant. It learns your patterns and saves you an hour a day.
Big congrats on launching! Really like how HeyHelp learns your patterns, drafts replies naturally, and keeps the inbox clutter-free. Looks like a huge time-saver!
@yasmin_baars_joras indeed! If you need any help setting it up or while using it, we'll be around to answer any questions :)
@yasmin_baars_joras I believe you'll save even more time when we release the next things that we have planned on the roadmap! AI follow ups, Custom Tags, and AI calendar scheduling, and so many others!
been using this for a few months now glad that I made the investment. Great product, attentive dev team. keep going!
HeyHelp
haha love this! Thanks for sharing your experience here 🙏