DEV Community: Harsha Kumar

The skill I built for my product turned out to be sellable on its own

Harsha Kumar — Thu, 25 Jun 2026 04:53:01 +0000

For the last 4 months I've been building Xedge — helping people find the right AI tool stack for their specific goal.
This week, needing to cover some short-term costs, I realized something: I've been doing this exact service for free in DMs for months. Someone asks what they're building, I figure out their AI stack, I reply with the breakdown.
That's a sellable skill completely separate from the product itself.
So I'm testing it directly — listing "AI stack research and setup" as a freelance service while XEdge keeps compounding in the background.
The lesson here: sometimes the tool you build to solve a problem at scale has a manual, sellable version hiding inside it. Don't assume the only output of months of building is the product itself — the expertise you gained along the way often has its own market.
Anyone else found unexpected income inside a skill they built for something else entirely?

The validation prompt library that's saved me from building the wrong thing 3 times

Harsha Kumar — Wed, 24 Jun 2026 04:49:59 +0000

Every time I've skipped validation while building XEdge, I've paid for it later with wasted development time.
Here's the exact process I use now before building anything new:
Competitor gap analysis: Claude

Describe the idea, ask for the top 5 competitors and their weakest reviewed feature. The gap between what exists and what's missing is your opportunity — or proof there isn't one.
Real complaint language: Perplexity

Search for actual user complaints about existing solutions. Not marketing copy — Reddit threads, review sites, support forums. The language people actually use reveals the real pain point.
Demand estimation: Google Trends

Search volume for the problem, not your solution. People search for "how to find AI tools for X" far more than they search for "AI tool directory" — that's the keyword gap that shaped how XEdge is positioned now.
The objection list: Claude

Ask directly: "what would make someone NOT buy this even if they have the problem?" List 5 real reasons. Most founders only think about why people WOULD buy.
The waitlist test

Before building anything new, I test with a simple landing page and waitlist. If people won't sign up for something free, they won't pay for it built.
This process has saved me from building at least 3 features XEdge users never would have used.
Full prompt library at xedge.tech.
What's something you built that you wish you'd validated first?

Day 1 launch numbers and why I'm ignoring the upvote count today

Harsha Kumar — Tue, 23 Jun 2026 05:45:28 +0000

Yesterday XEdge launched on Peerlist's Launchpad.
13 upvotes. 1 comment thread. 3 pieces of direct feedback.
Here's an honest take on what each number actually means:
Upvotes are a popularity signal, not a quality signal.

They tell you people saw your launch and clicked one button. That's worth something for visibility, nothing for product direction.
Comments are a depth signal.

One real comment thread taught me more about how a stranger perceives XEdge than 13 silent upvotes combined.
Feedback is the only signal that compounds.

3 people took the time to tell me what's confusing or missing. That's 3 free product audits from people with zero obligation to give them.
Today I'm not checking the leaderboard again. I'm going through each piece of feedback line by line and deciding what changes this week.
The lesson: optimize launches for feedback quality, not vote quantity. The votes fade in a week. The feedback can fix things that matter for months.
xedge.tech — still building, still listening.

What changed between 3 users a day and 23 users a day

Harsha Kumar — Mon, 22 Jun 2026 04:17:15 +0000

Three days ago XEdge was averaging 3-5 users daily. Yesterday it hit 23 — the best single day in 4 months of building.
Nothing fundamentally changed about the product. Here's what actually shifted:
Direct outreach compounds differently than content

I spent the last week personally DMing people who'd engaged with my content over months — asking directly for support ahead of a Peerlist launch. That direct outreach drove a 1,100% increase in direct traffic. Passive content alone never produced numbers like this.
Organic social finally caught up

Months of consistent posting meant that when I had something concrete to share — a launch, a milestone — there was already an audience primed to respond. The content wasn't wasted; it was compounding quietly the whole time.
The product experience mattered once people arrived

Goal Stacks page views are now outpacing the homepage. People aren't bouncing — they're using the actual stack builder. That retention signal is what makes growth sustainable instead of a one-day spike.
The lesson: growth isn't usually one big thing. It's months of unglamorous consistency that suddenly looks like momentum once you give people a specific reason and a specific ask.
Launching on Peerlist today. We'll see where this goes.
xedge.tech

What 4 months of building taught me right before my first real launch

Harsha Kumar — Sun, 21 Jun 2026 04:51:58 +0000

Tomorrow I launch XEdge on Peerlist's Launchpad.
It's not my first time putting XEdge in front of people — I've been posting, building in public, talking to users for 4 months. But it's the first time there's an actual "launch day" with a ranking, a leaderboard, and a finite window where momentum either builds or doesn't.
Here's what I'm carrying into tomorrow:
The product matters less than the relationships built before it

The people who'll upvote tomorrow aren't doing it because the product is perfect. They're doing it because I showed up consistently for 4 months and asked them directly this week.
Launch day is a moment, not the whole story

If tomorrow goes well, it's a spike. If it goes quietly, the daily floor of 3-7 users I've built over the past month doesn't disappear. The compounding work continues either way.
Doing the unglamorous stuff matters

This week I spent a full day on Qapita signups and screenshot emails for an accelerator program instead of posting content. That trade was right. Not everything that matters is visible content.
Tomorrow's the test of 4 months of consistency. We'll see.
xedge.tech — launching on Peerlist tomorrow.

What I learned asking 20+ people for support before a product launch

Harsha Kumar — Sat, 20 Jun 2026 03:47:56 +0000

Most advice about "building an audience before you launch" is vague.
Here's what actually happened when I did it concretely.
This week, ahead of XEdge's Peerlist launch Monday, I personally DMed everyone who'd engaged with my content over the past 4 months — comments, replies, genuine conversations.
The message was simple: "Launching Monday, would mean a lot if you could upvote, here's the link when it's live."
Response rate: roughly 70% said yes immediately.
What surprised me wasn't the yes rate. It was realizing these people were already rooting for me — they just never had a direct, specific ask to act on.
Most builders assume their audience is passively watching and will act when the moment comes. That's wrong. People need a direct, low-effort, specific ask.
"Support my launch" is vague.

"Upvote this Monday, link coming, takes 10 seconds" is specific.
The lesson: relationships built through genuine engagement convert into real support when you ask directly and make it easy.
Launch is Monday. We'll see what 4 months of consistent showing up actually produces.
xedge.tech

Lessons from getting blocked by email verification for a week

Harsha Kumar — Fri, 19 Jun 2026 04:58:34 +0000

Spent the last week trying to get verified on Peerlist to launch on their Launchpad.
Hit every wall possible:

Google Workspace trial expired, no Gmail access.

GoDaddy removed free email forwarding.

Zoho and Google Workspace both pointed at the same domain, conflicting MX records.
Finally fixed it by checking DNS records directly, confirming Zoho was the active mail provider, and logging into Zoho Mail directly instead of assuming Gmail forwarding would work.
The lesson: when something "should be working" and isn't, check the actual configuration instead of assuming the obvious fix will work. DNS records don't lie even when your assumptions do.
XEdge launches on Peerlist Monday. Small technical win, big unlock.
xedge.tech

The content creator AI stack that actually works in 2026

Harsha Kumar — Thu, 18 Jun 2026 05:14:35 +0000

Content creation used to require either a team or years of skill development.
Now it requires a stack and consistency.
Here's the exact one I'd use starting from zero:
Format selection: Claude

Most people pick a format because it's trendy, not because it fits them. Ask Claude to evaluate your natural strengths — writing, speaking, visual thinking — and recommend a format based on that, not what's popular.
Content planning: Claude

The biggest blocker to consistency is the blank page every single day. Solve this once — give Claude 10 raw ideas in one sitting, have it expand each into a full outline. You now have a month of content from one hour of work.
Production: CapCut, Claude, ElevenLabs

CapCut handles video editing with AI-powered captions and cuts. Claude handles writing and scripts. ElevenLabs handles voiceovers if you don't want to be on camera. Production time drops by 70% compared to manual editing.
Visual identity: Canva AI

Thumbnails, templates, consistent branding — built in one afternoon, never thought about again.
Distribution: depth before breadth

The single biggest mistake new creators make is spreading across 5 platforms immediately. Pick one. Post daily for 90 days. Only then consider a second platform.
Total cost: $0

Time to first 1,000 followers: 60-90 days

Time to monetization: 90-120 days
Full content creator playbook at xedge.tech — every tool, every step, real timelines.
What format fits you best — writing, video, or audio?

What investors actually look for when you have zero revenue

Harsha Kumar — Wed, 17 Jun 2026 06:30:03 +0000

This week I had conversations with three different funding programs.
None of them asked about revenue first.
Here's what they actually focused on:
Growth efficiency, not growth size

500 users isn't a huge number. But 500 users with $0 marketing spend is a strong efficiency signal. They're not impressed by scale alone — they're impressed by what you achieved with nothing.
Consistency of execution

Every conversation referenced my content, my building-in-public posts, my visible activity. Investors can see effort now in a way they couldn't 5 years ago. Public building is now a diligence tool.
Honesty about the gaps

I told every program exactly where I'm weak — no revenue, no team, no formal incorporation. Nobody penalized the honesty. They used it to calibrate what stage I'm actually at.
A specific, defensible vision

Not "we're disrupting AI" but a specific articulation of the gap — discovery and execution layer for AI tools, a category nobody owns yet.
The lesson: pre-revenue doesn't mean unfundable. It means you need a different kind of proof. Traction efficiency, visible consistency, and honest specificity matter more at this stage than your bank balance.
Still building. Still figuring this out in real time.
xedge.tech

What happened when I submitted an honest VC application as an 18 year old pre-revenue founder

Harsha Kumar — Tue, 16 Jun 2026 04:13:40 +0000

Last week I applied to LvlUp Ventures.
I didn't have revenue. I didn't have a co-founder. I didn't have a network. I didn't have a warm intro.
I had 500 users, a real product, an honest story, and a pitch deck I built in one afternoon.
Here's what happened within 48 hours:
LvlUp Ventures selected me for their Power of the Pitch event — virtual pitch in front of Alumni Ventures, Insight Partners, Antler, Fidelity, and others. 1 in 3 companies historically receive an investment offer.
LvlUp Labs invited me to their weekly accelerator info session with their General Partner directly.
SuperCharger Ventures introduced me to their startup scout for Malta Cohort 7.0 — a 12-week in-person accelerator with up to €1.5M in non-dilutive government funding, fully covered expenses, and a €250K direct investment track.
I'm 18. Still in college in India. Pre-revenue. Solo founder.
None of this came from connections. None of it came from warm intros. It came from building something real, telling the truth about the numbers, and submitting one application.
The honest lesson:
Most founders don't apply because they think they're not ready. The investors looking for pre-seed and pre-revenue founders are actively looking for people who are building something real — not people who are waiting until they're perfect.
Your traction doesn't have to be revenue. It can be users, engagement, content reach, or just a compelling honest story about why this problem matters and why you're the person to solve it.
Apply anyway. The worst they can say is no.
xedge.tech — still building, still going.

What a day of personalized AI stack replies taught me about what people actually need

Harsha Kumar — Mon, 15 Jun 2026 05:04:42 +0000

Yesterday I did something different.
Instead of posting about XEdge — I asked people to tell me what they're building and replied with their exact AI stack.
Here's what I learned from 20+ conversations:
Everyone is building something different.

SaaS products. Newsletters. Freelance agencies. AI models. E-commerce stores. Content businesses.
Everyone has the same three problems.

They don't know which tools to use.

They don't know in what order.

They don't know if what they're building is actually worth building.
The non-technical founders are the most underserved.

One person DMed me saying she's "honestly really lost in the digital space" but willing to learn. She has a real business. She just doesn't know where to start with AI.
That person is not a niche. That person is the majority.
Most AI tool content is written for developers. Most AI tool directories are built for people who already know what they're looking for.
XEdge is supposed to be different — describe your goal, get your exact stack, regardless of your technical level.
Yesterday reminded me why I built it in the first place.
If you're building something and want your exact AI stack — drop it in the comments. Still doing this today.
xedge.tech

The no-code SaaS stack that actually works in 2026

Harsha Kumar — Sat, 13 Jun 2026 04:54:47 +0000

I've been building execution stacks for 15 different business models on XEdge.
No-code SaaS consistently surprises founders because the barrier is genuinely gone now.
Here's what the stack actually looks like:
Validation: Claude

Before touching Bubble — describe your idea to Claude and ask it to steelman the 5 biggest reasons it fails. This conversation alone has saved founders weeks of building the wrong thing.
Building: Bubble.io

Real database. Real authentication. Real workflows. No code. The free tier is generous enough to validate and launch a basic version. Don't pay until you have paying users.
AI features: OpenAI API via Bubble plugin

This is the unlock most no-code founders miss. You can connect GPT directly to your Bubble app in 20 minutes. Suddenly your no-code tool has AI built in and your positioning shifts from "simple tool" to "AI-powered platform."
Payments: Stripe + Bubble

Subscription billing in one afternoon. Charge from day one — even if it's $5/month. Free users tell you nothing. Paying users tell you everything.
Distribution: community first

Post in 5 communities where your target user already hangs out. Don't pitch. Find a thread where someone is complaining about the exact problem your SaaS solves. Solve it in the comment. Link naturally.
Total cost: $0

Time to live product: one weekend

Time to first paying user: 2-4 weeks if you execute
Full no-code SaaS playbook at xedge.tech — every tool, every step, real timelines.
What are you building?