I'm a former 57-year-old construction site manager. I built an Email Decision OS because I couldn't afford to miss another deal.

The "built for a problem I personally lived" origin story usually produces the strongest product clarity - the use case specificity comes through clearly here.

Construction + compliance is one of the more regulated industries in the US. OSHA rulemaking, state-level contractor licensing, prevailing wage updates - there's a constant flow of regulatory changes that most companies track manually or miss entirely.

If email is the core workflow for decision routing, there's a second layer worth considering: upstream alerts before the email decision even needs to happen. That's the goffer.ai angle - legislative webhooks that fire before the compliance email backlog appears. Different layer than what you built, but they compound well.

the bit about missing 3 days and getting buried under 100 emails really landed, anyone who's done sales remembers that exact dread. curious why you went telegram bot instead of a standalone app? feels like a smart way to skip install friction, but idk if people trust a bot with their inbox early on. how's that been going so far?

The part that makes this work isn't the AI triage, everyone has the same models now. It's that you actually understand a specific person: the one who lives on their phone between shifts and treats a full inbox as dread, not a workspace. That's a real wedge. Field sales, trades, site managers, anyone not chained to a desk is badly served by inbox tools built for knowledge workers at two monitors. Delivering through Telegram instead of yet another app is the right instinct for that crowd.

One caution from years of running a sales org: be careful leaning on auto-drafted replies for deal email too early. Triage I can trust is worth paying for on day one. A drafted reply that's 90 percent right can still cost a deal, so the trust bar there is much higher. Nail the classification and make it feel almost magic before you ask people to lean on Approve for revenue conversations.

57 and just getting started is the right energy. Build less, ship faster, and own the niche nobody big bothers to serve.

Build on brother!!! 55 YO here!!!!

"I deeply understood one problem. One product solved it." — This is the whole game right here. Not CS degrees, not YC, not fancy stacks. Just knowing the pain firsthand and building the solution yourself.

I'm on a similar path — solo founder, non-technical background, built an AI tool because I hated writing weekly reports every Friday. Still early, still shipping. Stories like this remind me why we do this.

Wish you the best finding your first 100 users. Count me as one who's rooting for you.

One practical onboarding idea: make the first run read-only and show the user 10 real emails with your labels plus a short reason for each label. Let them correct a few before asking them to approve drafts.

For this kind of product, I think trust will come less from saying "AI handles email" and more from proving that the triage matches how they already think. If the first session feels like calibration instead of delegation, Gmail access and approval anxiety both get smaller.

I'm building something similar in spirit, just a different domain: a behavioral tool. The insight that got me started: most of the mistakes we make, especially the ones we already know are mistakes, happen because we're acting, not thinking. The reactive state overrides the part of us that knows better. So my whole approach is to put a stop between the impulse and the action by building just enough friction to buy back a moment of thinking before the mistake gets made.
Reading yours, I think you're doing the same thing from the other side: your tool removes the noise so the only thing left is the judgment. Mine inserts a pause so the judgment can happen at all. Different mechanism, same enemy : the human acting on autopilot.
Curious how you handle the trust step — letting an AI pre-judge your inbox is a big ask. How are you getting people over that first hurdle of handing the judgment to the tool?

ngl the bit about learning AI on your phone between shifts in yeosu is what stuck with me. wild path from IT sales to tile floors to building this.

"This is inspiring. Built between construction shifts and late nights—you embody the true spirit of a solo founder. Age is just a number, and your drive is incredible. 'Slash it' looks highly practical. Keep pushing, Min chul!"

The most valuable line in here isn't your age, it's "8 years in IT sales, 30 emails a day, miss 3 days and deals are gone." That is the moat. You are not building an email tool, you are building "never lose a deal in your inbox again," and you lived the pain, which most founders chasing this space never have. Two things from someone who has sold software for a long time. First, narrow your first 100. "People who drown in email" is everyone, which means no one. Go straight at solo founders, salespeople, and agency owners where one missed email is a lost contract, because that is your own story and they will feel it instantly. Second, a bot that reads someone's Gmail lives or dies on trust. Put the permissions and what you do not store front and center, or cautious buyers bounce before they try it. The story is strong. Point it at the people who bleed money from a missed email.

Really like how the email classification handles triage, most tools stop at sorting but Slash It actually drafts replies too. One thing I've noticed with pairing AI triage and dictation is that once the AI flags what needs a reply, the bottleneck shifts from "which email matters" to "actually writing the reply fast enough." I built DictaFlow for exactly that, hold a hotkey, speak the reply, release, and it types wherever your cursor is. Cuts reply time from a few minutes to a few seconds for short emails. Curious if you've thought about dictation as the input layer on top of the classification.

What a fascinating journey. Your determination to pivot from construction to building an Email Decision OS is truly inspiring, especially given the challenges you faced. I’d love to hear more about the specific pain points you encountered while using email in your previous IT sales role. How did those experiences shape the features you prioritized in your new tool?

Great example of solving a problem that actually matters in real life, not just another SaaS idea. Respect the focus on real outcomes.

This sounds like a very real problem, especially for people who manage deals, clients, vendors, or long email threads.

Email becomes messy because it mixes everything together: decisions, follow-ups, context, objections, documents, and small commitments. The hard part is not just replying faster, but knowing what matters, what is waiting on someone, and what could be missed.

A useful email workflow should probably help separate information into clear buckets: urgent decisions, pending replies, promised next steps, risks, and context for later.

That is where automation can be valuable — not by replacing judgment, but by making sure important signals don’t disappear inside a busy inbox.

57 and shipping a live product built between construction shifts, that's a better founder story than most of what gets celebrated here. And the "no inbox, just judgment, you tap approve" model is the right call , keeping a human on the approve button is exactly how these tools earn trust. Rooting for you finding your 100!!!

This is genuinely inspiring. What stands out isn't just the product, but the fact that you built it from a problem you personally experienced and understood for years. Many founders chase trends, but you've focused on solving a real pain point with practical experience behind it. Wishing you success with Slash—it’s a great example that innovation and entrepreneurship have no age limit.

This really resonates — I'm a construction PM who just built two AI apps with zero coding background. The 'built it because I needed it' motivation is so much stronger than building something because you think it'll make money. How long did it take you to get your first paying customer?

the red yellow white classification system is simple enough to actually use which is the right call. but curious how the AI decides what's urgent without knowing the context of your specific business relationships. an email from a client you've been chasing for six months reads differently than one from the same company if you just closed the deal. does the system learn your relationship context over time or is the classification based purely on the email content itself

What makes this interesting isn't the AI or Telegram part. It's that you spent years living with the problem.
A lot of people build solutions for pain they've researched. You built one for pain you've personally experienced. That usually leads to much better products.

Really inspiring story. Building a product at 57 while working full-time in construction is impressive. The focus on reducing decision fatigue instead of just managing email is an interesting angle. Wishing you success with the founding user launch.

The Telegram layer is smart because it avoids asking busy people to adopt another inbox. The place I’d be most careful is trust calibration: for the first 20 users, I’d track every “handle now” decision and ask which ones were false alarms or misses. That gives you safer onboarding and much sharper positioning at the same time: “we catch the revenue-risk emails you would otherwise miss.”

Sales, recruiting, and owner-operators feel like the right first wedge because the cost of one missed message is obvious.

This is inspiring. Most people think coding is the hardest part but stories like this prove that the real challenge is identifying a real problem worth solving. Building something out of genuine frustration is the best validation. Following to see how this grows.

The origin story is the pitch honestly. Someone who lost real deals to buried emails built the fix. That sells itself. Sales reps and recruiters are probably your fastest first users. One saved deal makes 39 dollars feel like nothing to them

Ok

This is a great example of solving a real problem from personal experience. Building an Email Decision OS after years in construction shows that valuable startup ideas can come from any industry, not just tech. I like the focus on preventing missed opportunities—pain-driven products often resonate the most with users. Wishing you success as you continue refining and growing it.

8 years in IT sales before construction — that's a real origin story for an inbox tool. People who've lived through the chaos of missing a deal because of a buried email build very different products to those who just read about the problem.

The Telegram delivery is smart. People overwhelmed by email are usually also overwhelmed by another app. Meeting them where they already are removes one friction point before they even start.

Quick question: are you seeing more demand for the triage classification or the draft-reply feature? I'm curious which one people actually use vs which one sells the sign-up.

I'm just getting started.

thats the spirit!!!!

Strong founder story. The one thing I would test early is how you answer the privacy/trust objection.

Because this touches Gmail, AI, and Telegram, the buyer who has expensive emails is also likely to ask: what exactly gets read, where is it processed, what is retained, and can I revoke access cleanly?

I would make that part as concrete as the product pitch. Even a short “how we handle your email data” section with plain-language limits would help. For a tool that decides what matters, trust is part of the product, not just a legal footer.

Min chul, this is incredibly inspiring. "Built between construction sites and late nights" — that line gives me goosebumps. As a fellow indie hacker building in the trenches, your story is the perfect reminder of why we do this.

Interestingly, I recently built something very similar using Make and gpt-4o-mini to solve my own friction. My setup automatically watches incoming job emails, filters them, and drafts highly tailored pitches straight into Google Sheets to save pitching time. Seeing how you beautifully packaged a similar automation flow into an "Email Decision OS" via Telegram and priced it at $39/mo is a huge eye-opener for me.

Huge respect to you for launching at 57. You're proof that it’s never too late to build. Cheering for your first 100 founding users from Singapore!

"This hit close to home. I'm also building as a
non-developer in Korea — NDT engineer by day,
solo app builder at night. 62 days in, just shipped
my first real product today. Your story is the kind
that makes this journey feel less lonely.
Rooting for you."

Hi Min Chul Choi,

First of all as a senior developer I find it great that you are still at it at the age of 57. I wouldn't say congrats, because I know people who have retired as developers at the age of 73, so there is still some years to go.

I looked your product, and straight of the bat I can see this being useful in sales. You might have some competition from CRM systems though. I got some ideas for you..

The name "Slashit". No offense, and this is highly subjective off course, but I dont recognize the functionality in the name. I did some research for you and found a catchy name that might be better and that is available as an .io, .ai and .tech domain. What about mailservant.ai - wouldn't that draw attention?

The integration. Again very subjective so bear with me - I've never used Telegram, and I don't know anybody of my friends who use it either (I'm located in western Europe). What about integrating this functionality as a Chrome Extension instead? Everybody uses Chrome these days.

Anyways, good luck with your product, I hope you are successful.

The strongest thing you have is not the Telegram bot, it is that you lived the exact pain. You lost deals because an email got buried. Most founders build for a problem they read about, you built for one that cost you money. Lead with that story everywhere, not with "Email Decision OS." The abstract name hides the part that actually sells.

The real insight here is that you kill the inbox instead of organizing it. "Stop opening your email" is a sharper promise than "classify your email." Worth testing that messaging.

For your first 100, skip generic founder channels and go where a missed email costs real money: sales reps, recruiters, brokers, agency owners. One saved deal dwarfs $39, so the pitch writes itself. I started in IT sales doing 30 emails a day, so this one hit home. Rooting for you.

This story is strong because the product is not coming from theory. It comes from a real missed-deal pain.

The part I’d be careful with is the category wording.

“Email Decision OS” sounds smart, but the buyer may not wake up wanting an OS. They wake up afraid of missing the one email that costs them money, a client, or a deadline.

That is probably the sharper founding-user angle: not inbox productivity, but deal-protection for busy people who cannot afford to read everything.

Happy to put the tighter first-100-user angle in writing if useful. I think Slash it needs a very specific buyer frame before you push too hard for founding users.