Shiva Kumar Esakki Pandiyan

🛠️ Building and in public.
🚀 Solo founder, gap year.

[Shiva Kumar Esakki Pandiyan @shivae372]

In January 2025, right before my ISC boards that February, I was still spending most of my time on the Class 11 JEE syllabus instead of board prep.

I'd skipped basically all my pre-finals at school for JEE prep.

My attendance had gotten so low that my teachers were genuinely worried I might not even pass ISC 12.

Every day was pretty much the same.

7 km cycling from home to school. 6 km from school to JEE coaching. 7 km back home.
I'd get home around 9:30pm, eat, and then study until 12:30 or 1am.
Boards and JEE. Both.

The coaching itself barely helped. Most of what got me through JEE came from those late-night self-study hours.

And honestly, I hated how JEE made me learn.

Memorize a formula. Drill the same type of problem. Repeat.

But I did it anyway.

That's what everyone around me was doing, and I didn't really stop to question it.
Somehow, despite skipping all my pre-finals and spending most of my time preparing for JEE, I ended up topping ISC 12 in my school.
I cleared JEE Mains too, and started preparing for Advanced.

Then one night, I watched a YouTube video that made me seriously question the path I was on.

I'm not going to say what was in the video, or what I ultimately decided about Advanced.

What it did was make me stop running on autopilot.

I started thinking about what I actually wanted from college.

I didn't just want an engineering degree because that was the next step everyone expected.

I wanted something more practical. Somewhere I could actually build things, work on projects, experiment, and eventually have the skills to build something of my own.

I started looking at options outside the usual path as well.

Around the same time, a separate problem showed up.

There was a mismatch in my name across my educational documents.

It became a problem while sorting out my passport and other official documents, and fixing it pushed my college plans back.

So I ended up with a gap year.

At first, honestly, I had no idea what I was going to do with it.

My parents didn't really understand it either.

To them, I'd worked so hard for JEE and then suddenly had a year with nothing happening.

And I understood why that worried them.

There's also the whole log kya kahenge thing in Indian households -- what will people say?

I could feel that pressure too.

But if I was going to have the year anyway, I didn't want to spend it waiting.

I wanted to use it to get closer to what I actually want to do:

Build a startup.

And, honestly, I also wanted to make some money.

Not because I suddenly cared about making money for its own sake, but because I want to fund at least part of my future education myself.

I don't know yet exactly where I'll study, but I know I want an education that is practical and gives me room to build. Being able to pay for some of that myself matters to me.

I've been obsessed with electronics since I was a kid.

In Class 7, I built my first PCB in EasyEDA — a very simple audio amplifier.

Around the same time, I was building projects with Arduino and ESP32, and I took a year-long robotics course where I learned C properly for the first time.

I'd sit and read datasheets even when I barely understood what I was reading.

But my parents wanted school studies to come first, and at some point the pressure won. I scaled it all back for a while.

That interest never really left, even when I wasn't actively building.

But during the gap year, most of my time actually went somewhere else.

Software.

I realized that if I eventually want to build a serious hardware startup, I can't just understand hardware. I need to be able to build the software around it too.

I don't have a CS background, so I went extremely deep into AI-assisted coding.

Not because I wanted to "learn AI."

Because I wanted to be able to build.

I spent a huge amount of time figuring out how to make AI coding actually reliable - how to structure projects, give models the right context, debug, test, review, iterate, and keep things from falling apart as the codebase got bigger.

Eventually, I started building tools around the way I work.

I built claude-bootstrap, a system around Claude Code that handles things like project context, skills, agents, hooks, memory and self-improvement.

Then I built nodo, which maps a codebase and gives an AI coding agent the architectural context it needs.

Those weren't just side projects.

They became part of how I actually build.

The public versions of nodo and claude-bootstrap are only a fraction of what I use inside my own SaaS workflow. The open-source versions are the parts I was comfortable putting out there; the internal systems go much further.

That whole process changed how I think about coding.

I went from not having a CS background to being able to take an idea, build it, debug it, and actually ship it.

And eventually, I started shipping products.

That's how Starvo came about - a review management tool for local businesses, built solo.

And Viridel - a private, local-first budgeting app.

It's launching in a few days, and I'm not treating the launch as the end. I plan to keep maintaining it, fixing things, and adding features based on how people actually use it.

The same goes for Starvo. These aren't throwaway projects I built just to say I built something.

They're products I intend to keep improving.

Hopefully, they eventually make some money too.

Looking back, none of this was the plan.

The plan was:

JEE Advanced → engineering college → the usual path.

Instead, I questioned that path, got pushed into a gap year because of a document problem, and decided to use that year to build the skills, products and income I could carry into whatever comes next.

I still don't know if my parents completely understand it.

Maybe they won't for a while.

But I hope someday they look back at this year and don't see a year I wasted.

I hope they see the year I started building toward something of my own — and started figuring out how to pay for the road ahead.

[Shiva Kumar Esakki Pandiyan @shivae372]

Totally fair—price is only one axis; the day-to-day workflow matters more. I’m exploring a private local-first Ready to Assign alternative that avoids bank login and keeps the ledger on-device. VIP waitlist for Viridel’s €39 lifetime launch offer opens Sep 10 for 24h: https://t.co/KoonztAako

[Shiva Kumar Esakki Pandiyan @shivae372]

Actual is a strong example of the privacy/local-first end of the spectrum. The tradeoff I keep hearing is setup and manual import versus convenience; I’m exploring a private local-first Ready to Assign workflow that keeps the ledger on-device and avoids bank login. VIP waitlist for Viridel’s €39 lifetime launch offer opens Sep 10 for 24h: https://t.co/KoonztAako

[Shiva Kumar Esakki Pandiyan @shivae372]

Interesting comparison—the $1.50/mo hosting angle makes Actual more approachable, while the tradeoff is still self-hosting and bank-sync trust. Im exploring a local-first alternative for people who want YNAB-style envelopes without bank login; VIP waitlist for Viridel’s 39 lifetime launch offer opens Sep 10 for 24h: https://t.co/KoonztAako

[Shiva Kumar Esakki Pandiyan @shivae372]

@FuriousEconomy Exactly—the question isn’t whether screenshots can be imported; it’s whether the budget can stay correct without a third party holding the credentials. That’s why I’m exploring a local-first path: more user effort, but clearer boundaries.

[Shiva Kumar Esakki Pandiyan @shivae372]

@heynavtoor The interesting tradeoff is that the cheapest option can also be the most legible: local data, no credential handoff, and a clear exit. That’s the direction I’m taking Viridel—private, local-first budgeting. VIP waitlist until Sep 10 (€39 lifetime): https://t.co/MOEYzSu7fA

[Shiva Kumar Esakki Pandiyan @shivae372]

@AyanakojiLedger That social layer is the blind spot: real budgets are negotiated in context, not just categorized after the fact. Designing for those small asks could make money tools feel more human.

[Shiva Kumar Esakki Pandiyan @shivae372]

@budgeteerpro Trials are useful, but bank sync is also where trust and portability matter most. A private, local-first option that avoids bank login is a different tradeoff. Viridels €39 Lifetime VIP waitlist is open until Sep 10: https://t.co/MOEYzSu7fA

[Shiva Kumar Esakki Pandiyan @shivae372]

@rmcastil @bot @ynab Interesting experiment—pricing is part of the UX too. When recurring cost is the friction, a local-first alternative can make the tradeoff clearer. Viridel has a 39 Lifetime VIP waitlist until Sep 10: https://t.co/MOEYzSu7fA

[Shiva Kumar Esakki Pandiyan @shivae372]

@personal_guru_f The smallest recurring expense is often the easiest place to create momentum. Tiny cuts become meaningful when they stay visible.

[Shiva Kumar Esakki Pandiyan @shivae372]

@ricardo_mena @spendeeapp Bank sync is convenient when it works, but it shouldn’t be the only path. A local-first ledger gives you a fallback when the connection breaks. That’s the line I’m exploring with Viridel. VIP waitlist until Sep 10: €39 lifetime: https://t.co/MOEYzSu7fA

[Shiva Kumar Esakki Pandiyan @shivae372]

@BuildWithCausal @zoop_design Starting with less ceremony: local-first data and a calmer path from income to envelopes. I want the budget to stay useful even when the subscription question is hanging over it.

[Shiva Kumar Esakki Pandiyan @shivae372]

I keep coming back to the YNAB question: should a budgeting tool cost €100+ every year just to help you assign money? Viridel is local-first and private. VIP until Sep 10: €39 Lifetime. https://t.co/MOEYzSu7fA

[Shiva Kumar Esakki Pandiyan @shivae372]

@FuriousEconomy That’s the tradeoff I’m wrestling with: bank sync is convenient, but I don’t want a hosted budget app to become the only place my money data lives. Viridel is local-first. VIP waitlist until Sep 10 (€39 lifetime): https://t.co/MOEYzSu7fA

[Shiva Kumar Esakki Pandiyan @shivae372]

@mitchelllong43 @Plaid @grok @bot @ynab The comparison worth making isnt just what the bot can do, but what stays legible when it’s wrong. A visible source, reversible edits, and a clean handoff back to the YNAB ledger matter more than an impressive first pass.

[Shiva Kumar Esakki Pandiyan @shivae372]

@dmnplb @Flomerboy @copilotmoney @ynab @getquin_app Investment tracking is a different data model from envelope budgeting: positions, cost basis, and market value each need their own history. Treating them like ordinary categories is where the edges get blurry.

[Shiva Kumar Esakki Pandiyan @shivae372]

@BrutusVTI @tyruspradier Reverse budgeting is interesting because it changes the planning horizon without abandoning zero-based discipline. The spreadsheet preview can expose which categories need flexibility before the month starts.

[Shiva Kumar Esakki Pandiyan @shivae372]

@yjsoon @FiachraRM That last stretch—polish and UX—is where a budgeting tool earns trust. The model can be elegant, but if recurring tasks feel frictionless and corrections are obvious, people actually keep the habit.

[Shiva Kumar Esakki Pandiyan @shivae372]

@talk2aks Measuring first is the unglamorous product work: reliable transaction capture, clear categories, and a way to correct mistakes. Without that foundation, an AI prompt just produces confident summaries of missing data.

[Shiva Kumar Esakki Pandiyan @shivae372]

@UpHonestReal The hard part isn’t choosing among tools; it’s making one trustworthy workflow survive the handoff between them. For money especially, fewer boundaries and a clear source of truth beat another dashboard.