I haven't doomscrolled in a long time (probably 20-30 days now). I'm surprised myself. Can't remember the last time 😅
— Tejas Kalpande (@tejaskalpande) August 2, 2026
↳ Building in public
Building in public
My journey into entrepreneurship — posting my day-to-day wins, failures, learnings, and doubts as I live it.
It's for me to look back on later, but it's also for anyone following along or trying to do the same.
Building
01 From X
Touched grass yesterday after 10 days of grinding. Was feeling burnt out. Probably should do often. https://t.co/1PMKQOL7Vd
— Tejas Kalpande (@tejaskalpande) July 31, 2026
Built and shipped a working tool in four hours at the @SarvamAI Epoch Buildathon (hosted by @GrowthX_Club, at @Razorpay's office in Bengaluru).
Used Sarvam's speech-to-text models (saaras:v3, sarvam-translate:v1, sarvam-105b), strong on Indic languages and accents, to analyse https://t.co/JCjX8YmBA3…
— Tejas Kalpande (@tejaskalpande) July 29, 2026
Two of my greatest idols on the same cover page, Musk and Naval. I'm going to devour this!
— Tejas Kalpande (@tejaskalpande) July 28, 2026
Musk: Builds stuff others can only imagine.
Naval: Tells you stuff others can't even think of.
@EricJorgenson keeps on creating amazing things for humanity. He too deserves the place in https://t.co/bLhjMOsP0U…
Participating in the GrowthX × Sarvam Epoch Buildathon at Razorpay's Bengaluru office, building a production-grade product in 8 hours on Sarvam AI's Indic language stack (speech, translation, document intelligence) to solve real problems in India's massively multilingual economy. https://t.co/ErdeI5I6qY…
— Tejas Kalpande (@tejaskalpande) July 26, 2026
Attended Agent Summit BLR organized by Magicball community. Booths, demos and talks. Here are things I observed:
— Tejas Kalpande (@tejaskalpande) July 24, 2026
1. Products are not agents but for people building agents
Most of them were observability, evals, agent verification, guardrails against agents doing damage. The buyer…
Stepped into a corporate office today, felt like a prison. Hope I never have to be an employee again. Absolutely don't recommend reading @MJDeMarco's Millionaire Fastlane if you are a happy employee. It will break your rose-tinted denial glasses and make you unemployable for…
— Tejas Kalpande (@tejaskalpande) July 20, 2026
A few ways I could go about it:
— Tejas Kalpande (@tejaskalpande) July 18, 2026
1. Consult for businesses, providing AI automation for their workflows (I get to see broken workflows across businesses until a pattern emerges)
2. Find people venting about their tools on G2, Capterra, Reddit, and LinkedIn
3. Mom Test-style…
As a beginner entrepreneur, most of your ideas are B2C since you've been a consumer always.
I also had a few of them, I assumed I'd start with one. But as I learnt and grew, I put those on the back burner, since B2C isn't really suitable for bootstrapped founders due to low…
— Tejas Kalpande (@tejaskalpande) July 18, 2026
I've open-sourced the AI agent skills I use daily. Made them generic so any agent can use them. Point yours at one and ask you for whatever placeholder info it needs.
A few worth checking out:
— Tejas Kalpande (@tejaskalpande) July 17, 2026
🔹 obsidian-notes-from-images: dump in rough notes, typed or photos of handwriting,…
I shipped ScreenMate, a local desktop AI assistant that watches your screen and answers questions about it. Electron + React + Ollama (Gemma4:e2b). No cloud calls, no telemetry, nothing leaving your machine.
https://t.co/y95Ku51xd7I had developed it as learning project. Plan to…
— Tejas Kalpande (@tejaskalpande) July 13, 2026
My current focus is finding and validating ideas and learning about GTM and distribution. I can build and scale but what might hinder me is GTM (Very new to this).
Various ways of GTM -
→ Relationship-led: You get customers with warm outreach within your network.
→ Sales/outbound/founder-led: Reach out to people/companies, pitch, show value, close.
→ Product-led: User tries your product, finds value, buys.
→ Marketing/content-led: You publish, content/SEO pulls people in.
→ Community-led: Build trust inside a community, community trusts you enough to buy.
→ Partner/channel-led: agencies or resellers refer you to their customers for a cut.
— Tejas Kalpande (@tejaskalpande) July 10, 2026
A few months ago I quit my job to pursue entrepreneurship full-time. Here's how it's going:
My life felt empty, like there was more to it than this. Realised entrepreneurship is how I'll be able to build a life of adventure, freedom, innovation, and meaningful contribution to society.
Quit the job, and for the first few months, learnt and built with recent tech like full-stack, AI, agents, and automation.
Realised I was a mediocre salaryman who didn't know the B in business. So put in the time to learn it: Books like Anything You Want (@sivers), The Millionaire Fastlane (@MJDeMarco), Almanack of Naval Ravikant (@EricJorgenson), etc. Events and workshops in Bengaluru. Videos of @naval, @AlexHormozi, @danmartell, and indie hackers.
Struggled with a scarcity mindset and started adopting an abundance one.
Have started on a few projects. Generic, probably won't make money. Shipping and putting them out anyway.
Finally getting the hang of business (a little bit, still a long way to go).
Working on finding and validating ideas I can build and sell. The builder-to-founder journey.
Still figuring it out. No idea where this road leads. Follow along if you want.
— Tejas Kalpande (@tejaskalpande) July 9, 2026
I've never understood "Ideas are worthless, execution matters" or "Ideas are multipliers of execution" till now.
Realisation came from @MJDeMarco's Millionaire Fastlane.
You cannot expect to have a unique idea that 7 billion people on earth haven't already thought of. Almost all possible ideas have already been thought of and probably put into action too! Whatever cool idea you thought of already exists.
But you don't need a legendary idea too. You just need to find an unmet or poorly met need and solve it. Take something that exists and execute on it better. Improve it or spin it in a new direction. And you're in business!
— Tejas Kalpande (@tejaskalpande) July 9, 2026
Mentored at @AIBoomi 24hr buildathon Recently. Some things that stuck:
→ Scope ruthlessly. One persona, one problem, ship that completely. "It's for everyone" = built for no one.
→ Builder ≠ founder. A builder makes the thing. A founder makes it, sells it, builds a business around it.
→ Validate before you build @BalarNeel . Opinions are worthless, behavior is data. Stop asking "would you use this". Ask "what's your current process, what does it cost you, when did this last go wrong."
→ Explain your product in one sentence that anyone in the room can understand @smondal1008 . Find hero usecase, state tangible value.
→ Aim higher @avinashraghava . Don't leave with just a weekend project. Leave with a cheque from VC, a customer, or a new teammate. Building is easy, chase GTM. Build to a global standard.
Best feedback I got from a builder: "Tejas gave peak advice to narrow my scope, otherwise I would have been fumbling for 23.5 of the 24 hours."
Narrow the scope. Ship the one thing that matters.
#AIBoomi #StartupWeekend #BuildInPublic
— Tejas Kalpande (@tejaskalpande) July 2, 2026
Went to @aiweekendsxyz event yesterday where we came together to build our portfolios. Here's how I did it:
→ Created a project in Claude Cowork.
→ Asked it to draft a bare-bones structure for my portfolio site. Since it already knew my indie-founder background, it gave me a good starting template.
→ Had it break that structure down into design guidelines + content.
→ Reviewed the content and filled the gaps.
→ Asked it to turn everything into a complete spec document for Claude Code. Till here it took me 2 hours.
→ Fed that spec to Claude Code with instructions:
Use the frontend-design skill for the design
Use stop-slop skill for writing
Support dark + light mode
Make it mobile responsive
Run the main agent as the orchestrator and spin up sub-agents to delegate the the tasks→ Claude code built and deployed the site in 30 minutes.
A few tweaks are still needed. Will do it over next few days. Here's the link: tejaskalpande.com
Big thanks to @Saul1545301 for the nudge to add a /now page. It's done too: tejaskalpande.com/now
— Tejas Kalpande (@tejaskalpande) June 29, 2026
What is calculated risk exactly? I realised it when I read MJ DeMarco's The Millionaire Fastlane.
Calculated/Intelligent risk is risk with unlimited upside and limited downside. Moronic risk is where the upside is limited and the downside is a bottomless pit.
Walking away from job to start a business is an intelligent risk, since the upside is unlimited (wealth and impact) and the downside is that you fail and go back to a job. Racing your car is a moronic risk, since the upside is a temporary high and the downside is a horrible death.
Without this framework, you'll find yourself taking moronic risks instinctively and shying away from life-changing intelligent risks.
— Tejas Kalpande (@tejaskalpande) June 23, 2026
I'll be mentoring at @AIBoomi Startup Weekend, a 24-hour buildathon where small teams take a sharp idea and turn it into a working AI MVP, built from scratch on the floor.
June 20–21, in Bangalore.
I'm a software engineer and indie builder, with a background in engineering across networking, hardware-software design, and systems engineering, now building in AI, automation, and full-stack apps. If you're attending, here's what I can help you with:
Scoping a vague idea down to an MVP you can actually finish in 24 hours
Shipping fast with AI tools like Claude Code, plus no-code/low-code where it makes sense
Agentic AI and n8n workflows to automate the repetitive parts
Boosting your productivity by automating manual or inefficient workflows so you spend the time on what actually matters
Building the actual product: web apps, dashboards, backends, integrations
Debugging and getting unstuck when the build or workflow stalls, which it will at some point
You don't need a co-founder or a polished idea to take part. Just an idea worth building and the willingness to ship.
Register solo or with a co-founder here: luma.com/StartupWeekend…
— Tejas Kalpande (@tejaskalpande) June 19, 2026