Atharva Jadhav
@ThinkAtharva
Followers
128
Following
2K
Media
37
Statuses
162
IIT Bombay · Building software with AI Learning fast by doing · Staying with problems longer Thinking in public
Joined November 2024
I’m not trying to look smart here. I build things. I get stuck. I sit with the problem longer than I should. Eventually it breaks. I use AI to move faster. I learn by doing, not by watching. Sharing the process as I go.
3
0
14
Wrapping up my SaaS MVP today. Shipping it tomorrow if all goes well. Also realized Gemini has been doing a lot of the heavy lifting for me. Feels underrated compared to how good it actually is.
2
0
4
Shipped this small addition to my MVP today. A rules and templates directory to speed up building. Nothing fancy yet, but it already feels useful. Testing it on myself first. Then a few users.
4
0
5
We pay for every AI tool out there. Claude. GPT. Cursor. Gemini. But has anyone actually used this macOS feature? I just noticed it today.
2
0
9
Building this today. It solves a real problem for me, so I’m testing it with a few users next. If it clicks, I’ll ship it. Feedback welcome.
3
0
10
I see many founders giving long free trials to get users. That’s fine, but it often shows you don’t fully value the product yet. I realized I don’t want to build that way. If I wouldn’t pay for it, I shouldn’t expect others to. Charging early forces clarity and better feedback.
3
0
7
I built a similar prototype SaaS to better understand how https://t.co/jEIUHoJWKy works from the backend. Recreating the flow helped me see the constraints and decisions more clearly. Sharing the image since it shows the output well (not me in the pic 😅).
1
0
4
I always feel good using my own product. Not because I’m hoping it’ll make 10k a month, but because I built it and it actually solves my problem. That feels like the real first step of building something meaningful.
2
1
7
Today was about staying consistent and keeping momentum. Work moved forward. Stayed active. Progress across a few fronts. From tomorrow, the focus shifts to hands-on work. AI projects and SaaS exploration, daily. Less talking. More building.
2
0
8
Saw this today and it sparked some curiosity. Not something I’m working on directly, but it highlighted a few things worth keeping in mind. Sharing for context. Back to building.
1
0
8
Tried https://t.co/jEIUHoJWKy by @notiiivy and ended up pretty impressed. The output quality is clean and polished. Feels well thought out. Now I’m just curious how this works under the hood. Solid SaaS. Respect.
1
0
6
Ran into a limitation while building. Realized I was treating the symptom, not the cause. Spent today learning the fundamentals behind it. The path forward is clearer now. Back to building.
1
0
8
I agree with this, but I also disagree a bit. Keeping the 8–5 or not isn’t really the deciding factor. I’ve seen this pattern in totally different fields too. Take UPSC for example. Lakhs prepare, very few become IAS/IPS. The difference usually isn’t effort or visibility. It’s
The dropout founder myth: - Quit your job - Build in public - Go viral or go broke The actual path: - Keep the 8-5 - Build at night - Ship something that works - Not just something that compiles Stability funds obsession 🫡
1
0
4
Love Xnapper or CleanShot? They are Mac-only downloads. I wanted a tool that works everywhere, so I built SnapPop in 24h. • Works on Windows, Mac, & Linux • No download (Runs in browser) • Studio-quality exports. Try it here: https://t.co/swxTfn3x7E
#buildinpublic
1
0
6
Watching “I failed 27 startups and made $1M as a solo developer” by @marclou again. Every time I watch his videos, something clicks back into place. Less noise. More clarity. Explore ideas. Build fast. Ship. Drop what drains you. Repeat. Back to small bets.
2
0
10
Today I learned something important while building a SaaS. Building fast with AI is easy. Building something that actually feels right is harder. When it doesn’t click, forcing it only drains energy. So I paused. Not quitting. Just respecting timing and taste.
5
0
9
4am lesson. Writing code isn’t the hard part. Shipping it is. Linking projects, fixing access issues, deploying functions, wiring APIs — that’s where most of the time actually goes. Things finally started clicking tonight. Sleeping now. Polishing tomorrow.
3
0
5
It’s 4am. Today wasn’t about features. It was about permissions, secrets, deployments, and figuring out why things should work but don’t. Learned more today than from weeks of tutorials. UI cleanup tomorrow.
1
0
5
I know ChatGPT can generate formulas, but half the time I’m scrolling a long answer or debugging which version is correct. So I made a tiny tool that just returns ONE formula and nothing else. It removes the chat layer completely. https://t.co/s4VQWeDx2b
0
0
3