Logbook

I watched a summary video of How to Win Friends and Influence People by Dale Carnegie. It's very good. I'm going to try to put that in action. I created a few motion graphics videos using Opus 5.5. Turned out to be nice. I will post the launch/demo video of the DecisionDecode project.

I need to talk to people to find out the problems and validate the idea that I have myself. But I guess nobody talks about how difficult it is going out of your comfort zone and talking to people. By default you just try to procrastinate and fill your time with learning and building just to avoid talking to people. I really need to get beyond this inertia and go out and talk to people.

I went to IndiaFOSS 2026 conference. It was fine but it felt like there were more students and less professionals. But there were a lot of good initiatives and good booths, including open software and social service related ones. My most important takeaway was that open source contribution is not just code but also documentation, involvement in discussions, designs, testing, support, issue triaging etc.

I continued working on problem discovery and I used Grok today to see what kind of businesses or startups I can build in system software. Looks like the default method is to start with service-based and do some paid contracts so that you learn and find the problems. But I had said no to service-based, so it suggested that you build on the pain point that you have seen in your career. But for that as well I have to talk to my current peers in the industry to see what's the current shape of that pain.

I see two ways now: First, start service-based offerings and take contracts. Second, talk to my known people in industry or any other people in industry that I find from events or LinkedIn and ask them what pains they face and build solutions for them. Major revelation: both of these ways involve talking with people, asking for their problems, and building solutions.

But I'm still not sure about system software because I have seen that field and feel like not going deep but instead staying in the field of AI and automation because it interests and excites me. Thinking more. One action item is certain: I have to seek people out, talk to them, and try to solve their problems. Currently I will not limit myself to one particular vertical but I will talk to people across domains and businesses. So, I will start ugly, start fast.

Another thing I could do is talk to people of various professions, non-technical ones. They might be drowning in a lot of manual work. We can build a solution which takes care of their profession holistically.

In the end, what is a business? Simply stated, a business helps people by solving their problems. If you just start helping people, you are in business! How to help people? By solving their problems. How to know their problems? By talking to them and genuinely caring about their work, needs and goals. And offer to help! Be useful to others.

When I ask myself how many people I have helped since I started on this journey, the answer is almost none, except just myself. So I'll get started with helping people.

Naval says you get rich by getting society what it wants but does not know how to get. You cannot get rich without producing value for society. You produce that value by solving society's problems. To know society's problems, you have to talk to the people in society.

I listed down the comprehensive summary of my complete work experience and skills. I discussed with Claude to find out what I am and what my actual skills are. Turns out I am actually a system software engineer, not an embedded engineer. This will help me finalize my domain if I want to do problem discovery in it.

Today I realized that when I was looking at some material from YC and articles from Paul Graham on getting ideas and problem discovery, they said that you should start with the problem, not the idea, and build a solution to fix the already existing problem. I realized that I also face a lot of problems on a daily basis, so instead of trying to come up with ideas, I will list the problems that I face, or that someone I know faces, or that the people I am talking with face. Something out of that could be a product. I also realized that some of my problems I can solve myself because I am an engineer and builder. So I can build a solution for myself. But the people who are not builders, non-technical people, might have the same problem. Then I can build the solution for them and make it a product. Have one problem like this, might build a plug-n-play product for it.

I talked with a friend today. He said that if you don't have a problem to build a solution for, you should probably go to SMB or tier 2 or tier 3 companies to understand their problems and build a solution for them. Doing this would give you an idea of what kind of problems exist in industry, or maybe you can build a product for a particular pain point. He told me not to restrict it to a particular domain or vertical; you can talk to businesses or industries of any domain or my own domain as well. I am thinking about this. Along with my product idea, I might also do this in parallel.

For my N-400 calculator, I have some Reddit threads to which I need to reply to promote my product. I will do that tomorrow.

You don't need an idea. You just need to decide on the domain that you'd want to spend the next few years on and see where the demand is. See the things that people are already paying for and for which there are already some companies offering the solution, and just do that. Competition is a sign of validated demand. Just get started because, anyway, when you talk to people, you will eventually pivot, so the starting idea doesn't matter.

Also, you should listen to your gut. Your gut tells you to do what looks hard and abstract. And your brain tells you to play it safe, small and structured. My gut tells me to build a full-fledged startup in the domain of my expertise, which is embedded systems. It will be painful. It will be hard and will demand a lot of me, which is exactly the opposite of what I thought when I started. I dreamt of having a one-person business, which I manage remotely and solo. This sounds nice and logical to me, but somehow it doesn't sit right with me. It feels like it is not who I am and I really don't know why I feel this. My gut says absolutely the opposite. So I'll start researching what the current demand is in this market. Talk to people and just get started with any idea. Any idea at all.

Was sick today, so couldn't work much. I posted on Reddit regarding my N-400 project. I will continue to do so for the next few days.

I realized that the problem that a green card holder has is: does the trip they are taking affect their N-400 filing date? If they have to do that in the tool currently, that means they have to enter all previous dates and then simulate the planning for the trip scenario. This is painful. I'm adding a save trips feature so that if they want they can have their trips saved. When they want to simulate a trip they can just log in to the app, add their trip, and see if it affects their filing date. I also added reminder emails when their filing date comes close.

I built my previous products just from my own ideas. That's building before selling, and it rarely works. You have to sell before you build. From now on I decided that I'll talk to people, figure out their pain points, try to provide a solution, and also check if that pain is big enough for them to pay.

Today I discussed with Claude the next steps in my journey. I checked with it:
- How can we do problem discovery?
- How can we talk to people?
- How can we approach them?
- How can we give something to them in return?
- What kind of mom-test-style questions can we ask?

I did some fixes to the N-400 Calculator project and published it on PeerPush and Product Hunt. Then I continued reading The Mom Test book. I had stopped reading for a while because of my dry eyes issue. Now I am recovering and able to read fairly.

The Mom Test book is also giving me ideas on problem discovery, which I am noting down as I go along. A real problem worth building a product for will cater to people's current focus and goals, something they are actively trying to fix or improve, and it should be a must-fix-now/painkiller kind of problem. Not nice-to-have/vitamin kind.

Then I published a launch post of my project on X and LinkedIn. I'm going to watch today's Grokbot live stream from SpaceX.

I wrote a guest post for the conference I had attended, AgentsNexus India 2026 in Bangalore. Well, I am no AI expert, but I wrote the article given where I stand. Tried to keep it honest and real. I hope readers like it. Here's the draft: Solid Ground in the Chaos of Agentic AI.

I attended an event on finding first 10 customers where the speaker was Pulkit Jain, co-founder of Vedantu. A really great session by Pulkit where he talked with us like a friend, not as an exec. His simplicity and grace carried this interactive and informative and fun session. Getting your first 10 customers boils down to what is unscalable: using your credentials, proximity, your network, or whatever is possible to get those first customers. Scaling and related techniques like ads, SEO, etc. come later.

The core learning for me, in addition to the business learnings, was that we must trust our gut over what you hear from other people or read from books and other sources. A lot of mental, emotional, unconscious calculations happen in the background and the gut feeling comes out. Listen to your gut and do that.

Also you must have a mission: a problem that you truly want to solve and firm belief in your way that the problem can be solved with this. Pulkit told a story that they were getting traction in the Middle East and the West with their products, but they did not proceed forward with that because their mission was to enable education for students from small cities of India with less access.

Full write-up: Finding your first 10 customers with Pulkit Jain of Vedantu

I added both of my new products Decision Decode and N-400 Calculator to my website and LinkedIn and I set up Google Analytics and Search Console for both of them, Google and Bing, with IndexNow. I also added some useful click counters via Cloudflare D1 so that I understand whether users are able to complete the full flow. If they start getting a fair set of users, I will add Google AdSense on them. Now I need to create more backlinks so that Google shows up my apps higher in search. For this, I will start making X, LinkedIn, Reddit posts and maybe other platforms like Product Hunt, Indie Hackers, etc.

Mark Lou dropped an ebook about his journey to $3M. I started reading it.

Today I worked on two of my shipped projects. I'm working on improving their accuracy and usefulness, and also making them SEO and AEO-friendly so that they get traffic. Fixed many bugs today, and I realised that I'm hitting my Claude Pro subscription limits often. I need to upgrade.

Also, I realised that entrepreneurship takes a heavier toll on your brain. When you're employed, your brain can rest once your task is complete. In entrepreneurship, ideas are your fuel, and they come asynchronously, when you're sleeping, taking a walk, resting, working out, basically any time of the day. When an idea comes, the brain starts spinning again, so my brain basically never shuts off, and this really exhausts me. I need to find a way to fix this.

I created Obsidian notes from my handwritten notes and agenda of AgentsNexus Conference 2026 using Claude and my self-hosted Obsidian Vault MCP server. Turned out nicely.

I also built the blog section of my website so that I can publish articles. The first article is from the notes that I created today. I'll also use articles as the means to expose my Obsidian notes publicly as needed.

I visited the AgentsNexus conference 2026 in Bangalore today. Yesterday it was workshops, where we learned about agentic security and hardening, and WebGPU and WebMCP. Today it was a full day of conferences where industry experts talked on various topics in agentic AI and LLMs. I liked the breadth of topics, where almost everything was covered — from LLMs, training, agents, memory, graphs, RAG, search, and all the current advancements and trends in the same. While listening to the experts and networking with people, I realised that what I know of AI and agents just scratches the surface, and its advances go very, very deep.

While sitting at the conference, I completed my development and deployment of one micro tool that I was working on. I used the remote control feature of Claude Code to access the session on my VPS. Really amazing that I can be anywhere and still ship something.

A few of my friends on the same path organised a co-build/hackathon session. We decided on a target and started building, and also made social media posts. I shipped one of my apps, which was pending for a long time. It's online and just needs a few tweaks and then I'll post on social media. I also created a small free tool, which I intend to ship and monetize using Google AdSense if it gets sufficient traffic in the future.

Today was a good day because I had the chance to get out of the house and interact with people, which I think I need to do more. I need to find some regular activity where I need to get out of the house. WFH is not working well for me (health issues). I was thinking of some part-time activity related to start-ups and tech, will explore.

My friends have a product in the pharma domain, and watching them work made me realise that you must have the domain expertise to understand the nitty-gritty stuff of it. This again pushed me towards building in my own domain expertise.

I posted on reddit regarding my YT like to view extension. Posted on Reddit for the first time. I also posted on X and LinkedIn regarding the MCP server addition to my self-hosted Obsidian knowledge base guide. Now it is a complete and actual second brain that all your AI apps and AI agents can work with. Today I was in dilemma whether I should create a SaaS product that I can operate solo or build a full‑fledged startup in the domain of my expertise, the embedded industry. I was under the impression that building a SaaS might be easier, but Claude told me that it's a misconception and it's not that easy to build a profitable SaaS any more than building a startup. I might take up Claude on that advice and start problem discovery in the embedded domain.

My like-to-view percentage browser extension for YouTube videos is approved by Chrome and Firefox stores. It's live. I made a post about it on my X and LinkedIn.

I am personally finding the extension useful and I realise that I use YouTube majorly on my phone, so I had created an Android app. Today I refined it and made it work. If I find it useful I'll publish it on the Play Store as well.

Today is a special day for me because I shipped a project in public for the first time! I know the project is very small but it's a step forward for me in building useful products. I don't have any big hopes for this but I hope it helps at least a few people. And I am onward on the journey of problem discovery.

Today I came to my alma mater, IISc Bangalore, to conduct mock interviews for the candidates of my department who will be appearing for campus placements this year. Feels nice to be back on the campus with amazing nature, brings back many memories.

The browser extension that I've built, I found it useful so I have posted it on the Chrome and Firefox extension stores. It's currently under review. Will update when it's approved.

I updated my self-hosted Obsidian AI knowledge base guide to include MCP server so any of your AI apps can access your knowledge base from anywhere. Guide is updated. Will post about it soon.

Turns out the Android app that I developed did not work in the first go. I am going to let Claude Code drive the phone with wired USB debugging and let it debug the issue.

As I had posted previously, I talk to random people every now and then. Today too, I had an interesting conversation which is kind of the general consensus around AI among the tech crowd.
They are mostly pessimistic about AI's impact on us, and I do see the practicality in it:
- People are losing jobs and are unable to cope with AI advancements
- AI is taking a downturn in the future with respect to the economy, environment, and the well-being of humans

But in contrast to the people I am talking with, I feel like I am majorly optimistic about AI's impact. I see AI as a revolution like previous revolutions, the industrial, computer, and internet revolutions. AI is going to be an extension to our intelligence, so that we express our creativity and give the job of materialising it into reality to AI.

What I hear often is "okay if AI is going to do everything, what are we going to do?"
To that I say: AI is going to elevate our goals as well. We are free to dream and aim as high as we can imagine, heights which were unimaginable before now. And to reach those goals we obviously need people, but with different expertise, because goals are being elevated to a new standard. The nature of jobs will change, but people will still be needed for performing them. When tech makes building stuff easier, we naturally aim higher to the next seemingly impossible goal. So I think the scenario where humans have nothing to do is not possible.

The harsh truth:We are unfortunately an eventually extinct civilization. We will never be able to escape our solar system and will be eventually wiped out by our own actions or by a cosmic accident.

But, a big but, with AI now solving decades and centuries-old unsolved math problems, there seems to be a very tiny hope that we might escape our eventual fate. What if AI enables us to discover new physics and solve the mysteries of the universe? We might survive beyond this planet and finally reach type 1 civilization on the Kardashev scale. Possibly reaching type 2 too!

Am I being overly optimistic here? Not sure. We'll see how it turns out, but some turmoil and downturns are inevitable with the hope of the generous upside that we saw with previous technological revolutions.

Today I created an Android app developed from the browser extension that was created yesterday. I'll start using it and if it's useful I'll release it on the Play Store.

For my self-hosted Obsidian AI project I added the custom MCP support so any of your chatbots like Claude and ChatGPT can access it using the MCP. With this, now you have a complete second brain. For me Claude is my personal assistant so it needs my knowledge base first-hand.

Someone starred my GitHub repo (Self-hosted Obsidian AI knowledge base) today! I know it isn't a big deal, but it's great knowing someone found my work useful.

Today I resumed work after many days and worked on a browser extension that I think would be useful for me. Opus 5 took an hour to build. Did a thorough job. Felt great when it worked. So awesome to know how amazing coding agents are. Thinking I'd post it on stores for others to use. It's a basic version, just gets the job done, would scale it if people find it useful.

I'm slowly getting back to work, hopefully I'll not burn myself out again. Will be restarting my X posts too. Good to be back. Dreadful journey of problem discovery starts again, haha.

Today I had a very useful conversation. I told people I'm in the problem discovery phase. I also asked the other queries that I have in this space and got the responses.

When I asked how I should do problem discovery, I got the following things from people:
1. Talk to people, code, security, marketing, etc, whatever you're interested in, or all, and then decide. Just talk to people and ask them their pain points. What would they pay for to get solved. LinkedIn, events, anything.
2. Check what YC and funds like Together Fund are funding, what problems they are funding, see if you're interested in those and iterate fast.
Do just to break fear or hesitation, then do because you want to do the thing, then decide you really want to do it or not.
In business, it's best to take something and go deep than go wide.
Go through YC Startup School course.
Try the T method. Try out multiple verticals and go deep. Go deep into one vertical and keep on talking to people, keep on digging, you'll find the pain points.
It's good to have something you're passionate about, or what your expertise is in. (Getting multiple signals here from multiple sources.)
3. You can see YC companies and just copy, it's fine, they fund the duplicate companies as well.
4. YC and funds and investors emphasize the founder more than the idea. So the founder must have some expertise or edge in the matter.
5. Co-founder should be from your network, who you know and trust deeply. They should challenge you, not agree blindly.

What I realised:
You should list down your expertise and edge.
Garry Tan also mentioned to stick to your expertise and passion instead of what's hot currently.

Public speaking, or speaking in general, is I think very important for entrepreneurship, so I've been practising my speaking, because whenever I go to the business events I see people who are very fluent in talking. I think our speech should be crisp, clear, smooth and non-hesitant, so I'm going to pick up random topics and talk about them for 5 to 10 minutes, and review them after the talking and have AI or Gemini evaluate it and give me suggestions.

Today I talked about a topic of financial freedom. I observed that I miss articles in places and I use too many filler words like umm, so, yeah, etc.

Today I took a good long rest, which I couldn't do over the weekend due to other commitments, and in the evening I went for a walk around the lake. I just struck up a conversation with a stranger and exchanged LinkedIn at the end of the conversation.

The one good thing that has happened to me since I started on this path is that as I am attending events and talking to people, I am getting better at talking to strangers. I feel less afraid and less hesitant to just walk up to a stranger and strike up a conversation. For me this is very new and this is something that I would never have thought was possible for me because I am an introvert by nature, but here I go. I am doing this and this is very rewarding. I mean the personal growth that I am seeing on this journey is very rewarding.

Today I watched Dan Martell's video on business, where he explained elaborately how to start a business from zero and scale it to a much bigger size. He explained:
- coming up with what you should do
- how to create the offer
- how to sell it
- how to talk to people
- how to scale your business

It was a very detailed talk and I realised that this is an all-new world that I didn't know anything about. I've been an employee and I have never seen a business working, so all these things bombarded at me by Dan Martell left me feeling afraid and excited at the same time, but I like it. That's the adventurous path I'm on.

Also today I read The Mom Test book, which is very insightful, and I'll be taking a break in the next few days to recover from my dry eyes. I'll post logs if I do anything interesting.

Went to the eye doctor today because my eyes were not recovering. Turns out I have severe dry eyes due to too much screen, so got some medicines and some precautions to take. Looks like I should work from outside the home, maybe cafes or coworking spaces, since you look outside of your screens, which relaxes your eyes.

Watched the video regarding the boring businesses. How boring businesses like software for pharma companies, even software for plumbers, are multi-million dollar valuation companies today. Extremely boring but painful, it's printing a lot of money today, so picking up a boring but painful niche is probably a good idea for a business.

And so I watched Dan Martell's video today and his life story, "Prison to a Multi-Millionaire," is proof that your excuses are not valid. Real OG guy.

So looks like my eyes are still not recovered, so I continued Sam Bhagwat's Principles of Building AI Agents book and later I watched Scale AI founder Alexandr Wang's interview. He talked about how AI is going to bring more human progress in a decade than the last century combined. He talked about loop engineering and designing feedback loops with which agents can work and interact with each other. He also said that the design always comes down to simpler stuff like markdown, skills, crons, nothing too fancy. He also mentioned that LinkedIn is for finding customers; technical writings on LinkedIn can be ignored. He also mentioned that as a founder you should have conviction about your idea and understanding of the future, otherwise you'll get drowned in the noise of others' opinions.

I also discussed with Claude voice AI regarding a few ideas that I had. It was an insightful conversation. Claude AI has certainly improved its voice capabilities, and during the conversation I realised that I was able to contribute to conversations effectively because of my experience of talking with people. I go to events and talk to founders of various companies. I know what the current trend of companies is. I was able to counter Claude's opinion and make it give me more accurate and relevant information. I realised that even if I don't have anything to talk about myself, I should still put my shame aside and talk to people at the events, so I'm going to do that.

Today was particularly hard for me emotionally. I went to a cybersecurity event today, which was nice. I learned many great things about AI attacking your applications and AI defending your applications. When I talked to other people I was low on confidence because I haven't built and shipped much and I don't have any product which I can sell. I don't even have the idea or the pain point I'm going to build a product for.

This harsh reality kind of hit me that after spending more than six months I don't have anything, but I did learn a lot of things in the past few months. Coming from an employee background, learning about entrepreneurship and business, a lot of misconceptions were cleared and it's a work in progress. I feel it's very slow. Also people recommended that I start a service business first so that it pays you and you learn along the way so that you don't have the feeling where you are not getting paid. I don't want to do any service-based business. I want to have a product that solves the real pain point and helps people.

Now I'm very very confused: should I really go towards a service-based business to learn about other businesses and their pain points and get a product idea from there, or continue working on my current strategy, which is going online or talking to people and finding out the real pain points in the industry and also looking out for problems in the AI space to solve?

I was a little down today but this is expected. I knew it wouldn't be easy when I started. I had told myself that this will not be easy. It will be painful so I'm feeling that pain now, but it's also fine and I'm growing. I think it's a good thing that you should have the goals, which we don't know how to reach. That makes us scared, anxious, and clueless.

Also I wish there was a guide to entrepreneurship, like a person who knows nothing about entrepreneurship, how that person can get started. Like a guide or roadmap which you can just follow through and just mark it down, like "learned this, did this, now this is the next step," kind of like that. It's wishful thinking but I wish something like that existed and if I am able to pull this off I'd probably like to do something like this: a book or guide which starts from scratch.

In the last few weeks I tried to work particularly super hard and I also tried to avoid instant gratification stuff and any kind of dopamine hit entertainment. I worked super hard and I feel like I did not rest enough. My mind was always filled and avoiding the instant gratification that we have the habit of took an immense amount of willpower. I got burnt out.

I took last week to rest and recover. My lesson here is that you can work super hard but then you also need to rest hard. You have to give your mind daily rest. You cannot engage your mind all your waking hours. Either that or my body will adapt to working super hard and willpower will increase so that I can push myself and avoid instant gratification stuff.

But I am still going to treat rest as a priority and I believe rest means anything that relaxes your mind, like:
- just sitting doing nothing or meditating
- just taking a walk around a block or garden
- any other things like that
- talking with your friends and family or cooking etc.

So I am going to try this out and update how it goes.

And so when I got burnt out, I also took this opportunity to do a dopamine detox. It means that I took the time to just sit and let myself be bored. I meditated maybe once or twice a day. I also tried to notice the feelings of boredom in me and let them stay there. I resisted the urge to pick up my phone or unlock it or watch something on it. I just let the boredom come to me and I just watched it.

This thing, I think, has worked out nicely. I observed that I was more present in the activities that I'm doing. My chattering of mind was also reduced. I was focused on the present moment without many future or past thoughts bugging me. I found peaceful moments peaceful, not disturbing or restless. I also started getting pleasure in the simple things like taking a walk, eating the regular food and finding it tasty and joyful, etc. My restlessness also reduced. Overall it was a great experience and I would recommend that you try it.