Forem: Blink

Day 9: When Streaks Break and Ambition Returns

Blink — Mon, 15 Dec 2025 19:59:43 +0000

The Streak Situation

Day 9 and I have to admit, the leetcode streak broke. Github streak too. Too many relatives around, schedule completely thrown off, couldn't maintain it.

But here's the thing -, I'm not pausing, just restarting. That's an important distinction. A broken streak doesn't mean you failed, it just means you reset the counter and keep going. The work still matters even if the number goes back to zero.

People get too attached to streaks sometimes. They become more about maintaining the number than actually doing meaningful work. If I miss a day because life happens, that doesn't erase everything I've learned or built. It just means Wednesday I start counting from one again.

Meeting Ambitious People Changes Things

Met someone recently who's really ambitious about what they're trying to build. It got me thinking, I should be pushing harder too.

It's easy to get comfortable with your current pace. You're doing something every day, so you feel productive enough. But then you meet someone who's operating at a different level and it reminds you that you could be doing more.

Not in a comparison way where you feel bad about yourself. More like they reminded me what focused ambition actually looks like. When someone's really locked in on their goals, it's contagious.

The Wednesday Reset

ML catch up begins Wednesday. Leetcode resumes then too. Fresh start, same energy.

I've been behind on my ML schedule for a bit now, but Wednesday is when that changes. No more "I'll catch up eventually" - actually sitting down and catching up.

Same with leetcode. Restarting the streak and this time being more realistic about maintaining it even when the environment changes.

Lets Be Ambitious

That's the energy for moving forward. Not just showing up and doing the minimum, actually being ambitious about what I'm trying to achieve here.

Day 9 done. Streaks broken, ambition renewed, Wednesday we reset everything.

Day 8: Grandma's Place

Blink — Sat, 13 Dec 2025 17:03:49 +0000

Early Morning (That Won't Last)

Woke up at 9 AM today. Sounds impressive but don't get too excited - I don't think that's continuing tomorrow. I'm back at my grandma's place and she's pretty old, so gotta visit. The schedule gets thrown off when you're not in your usual environment.

That's one of those things about consistency - it's easier when your routine is stable. Change locations and suddenly your 9 AM wake-up becomes noon again. But whatever, you adapt.

Scikit Learn Progress

Finished scikit learn sections 1.1.1, 1.1.2, 1.1.3, 1.1.5, and 1.1.15 today. Just to be clear - I didn't blast through all of these in one day. It took time to work through them properly, understand what each section was covering, actually absorb the information.

Too many people speed-run documentation and wonder why nothing sticks. I'm trying to avoid that trap. Tomorrow I'll work on the practical implementations of these concepts. Theory is one thing, actually coding it out is where you find out if you understood anything.

Stacks and Leetcode

Read about stacks today - another data structure to add to the collection. Did my leetcode problem as usual. Still maintaining that streak even though I considered dropping it yesterday.

Maybe I'll keep both going, maybe something will have to give eventually. For now I'm managing.

The Workout Constant

Hit the workout like always. That's probably the most consistent thing right now - no matter what else is chaotic, the workout happens. It's become non-negotiable at this point.

Things That Don't Work Out

Was gonna mention something else but yk, things usually don't work out the way I think they will. So not even gonna say it. Why jinx it.

That's day 8. Woke up early, probably temporary. Made progress on ML. Did the work. That's all there is to it.

Day 7: Trading Leetcode for ML (Temporarily)

Blink — Fri, 12 Dec 2025 18:53:52 +0000

Small Wins

Woke up at 11 AM instead of my usual 12. An hour earlier. Yeah I'm counting that as progress because when you're trying to fix your sleep schedule, you take what you can get.

The Leetcode Dilemma

Did a revision problem on leetcode today but I'm seriously considering dropping it for a bit. Here's the thing, I'm so behind on my ML schedule and I can't do everything perfectly. Something has to give.

Leetcode is important for interviews and problem-solving skills, sure. But right now ML is where I need to focus. Sometimes you gotta make strategic decisions about what to sacrifice temporarily. Maybe I'll come back to it, maybe the streak breaks. We'll see.

Logistic Regression and Documentation

Got into logistic regression today. Read through examples on the scikit learn documentation. I'm trying to actually understand what's happening under the hood instead of just copying code and hoping it works.

That's the trap with ML - it's easy to import a model, fit it to data, and call it done. But if you don't understand what logistic regression is actually doing, how it differs from linear regression, when to use it - you're just a code monkey, not someone who actually knows ML.

Thinking about checking out the California housing dataset question tomorrow. Or maybe not. Depends on how much I can catch up.

LinkedIn Posting Now?

Posted something on LinkedIn today: https://www.linkedin.com/posts/somay-kousis-630ab1313_smartindiahackathon-sih2025-teamwork-activity-7405310023767154689-p_iN?utm_source=social_share_send&utm_medium=member_desktop_web&rcm=ACoAAE-37UQBKy4jUXKIZLL50F1J2QdJ5Y6WyPU

Didn't think I'd become the type to post on LinkedIn but here we are. Building in public means actually being public about it, even on the corporate social network.

Arms and Dreading Leg Day

Trained arms today at the gym. Tomorrow is leg day and I'm already not looking forward to it. Leg days are always brutal but that's part of the deal. You can't just train the muscles you like and ignore the ones you don't.

Same principle applies to learning actually. Can't just do the fun parts and skip the fundamentals that feel boring.

Behind But Moving

Day 7 done. Behind schedule, making tough choices about priorities, still showing up.

Not every day is gonna be perfect. Sometimes you're just trying to minimize damage and keep momentum going.

Day 6: Rebuilding Plans

Blink — Thu, 11 Dec 2025 10:41:10 +0000

The ML Plan Needed a Reset

Scrapped my entire ML roadmap today. The first one looked good on paper but it wasn't clicking. I was forcing myself through content that didn't flow right, trying to follow a structure that worked for someone else but not for me.

So I rebuilt it. This time focusing on how I actually learn - what makes concepts stick, what order makes sense in my head. Generic roadmaps are useful as guides but you gotta customize them to your brain. Everyone processes information differently.

That's one thing about self-learning - you can be flexible. If something's not working, you can pivot immediately instead of being locked into a semester-long curriculum that doesn't fit.

Leetcode and Revision

Maintained my leetcode streak. Today was a revision problem - going back to questions I've already solved to make sure they actually stuck. That's the real test. Can you solve it again without looking at hints or your previous solution?

A lot of people rush through problems just to increase their count. But if you can't redo them a week later, did you really learn anything? I'm trying to balance moving forward with making sure the foundation is solid.

Core Work and Random Realizations

Skipped the gym today, did core work at home instead. I've never properly trained my abs before and it was harsh. Like genuinely brutal.

But it got me thinking - getting abs might actually be easier than building other muscles. If you train core consistently for a month, the results seem more straightforward compared to like, getting bigger shoulders or chest. Obviously it's not easy, but the path feels clearer. Just gotta commit and stop making excuses.

Random thought but sometimes workouts give you these weird moments of clarity about effort and results.

Mutiny Updates

The other founder was working on APIs today. Project's moving forward even when I'm focused on my own learning. Multiple things happening at once - code, ML, workouts, building mutiny.

Seeing if I can actually maintain all of it or if something breaks eventually.

That's It

Day 6 done. Rebuilt a plan, maintained a streak, destroyed my core, thought about abs.

Not every day needs to be eventful. Some days you just adjust course and keep moving.

Day 5: Catching Up

Blink — Wed, 10 Dec 2025 15:07:28 +0000

so here's how day 5 went

The Work Part

woke up late again, nothing new there. but hey, I maintained my leetcode streak so that's progress. soon as I opened my eyes, solved a question, then jumped into mathematics.

3blue1brown has been a game changer for me. like I genuinely thought matrices and determinants were just us messing around with random numbers for no reason, turns out there's actual logic and reasoning behind it? the way he explains things just clicks, you know. makes you wonder why school never taught it like this.

after that, watched some ML videos. being honest though, I'm still behind on my weekly plan. set these ambitious goals and now I'm playing catch up, but that's fine. by the end of this week I'll be back on track, that's the goal.

The Broken Promises

remember how yesterday I said I'd maintain github streaks like I do with leetcode? yeah about that... didn't happen. will try again from tomorrow. consistency is hard when you're trying to build multiple habits at once but whatever, we keep trying.

Gym and Other Random Stuff

so this temporary gym I have to deal with for 3 weeks is terrible. like really bad. I miss my good gym near college so much. after workout today I was riding back home and literally didn't realize how fast I was going till my song ended, that's when it hit me lmao.

home doesn't suit me. never has. I function better when I'm at college, away from all this. 3 more weeks of this then I'm back.

Where I'm At

leetcode streak: alive

github streak: dead (for now)

ML progress: behind but catching up

mental state: managing

location: wrong one

that's day 5. tomorrow we try again.

Day 4: Four Days of Work, One Strange Dream, and a Late Start

Blink — Tue, 09 Dec 2025 09:12:15 +0000

I woke up at 12 p.m. today. Not ideal, not tragic. Just late.
But waking up late didn’t cancel the last four days, so let me put everything on the table properly.

The list below isn’t “today’s grind.”
It’s the total accumulation of everything from Day 1 to Day 4.
A snapshot of where I actually stand.

The Four-Day Stack (Everything Done So Far)

DSA / Leetcode

Across these four days, I worked through the following problems:

217, 242, 347, 36, 128, 217 (again), 238, 125, 167, 15, 704, 74, 153, 121, 11.

Some were clean solves.
Some were messy.
Some needed peeking at solutions.

But it’s work done regardless—and I’m counting that.
Because honesty matters more than pretending I’m grinding flawlessly.

Math + ML Foundation

I went through:

Chapter 1 video
Chapter 2 video
Chapter 3 video
Chapter 4 video

All short, simple, digestible—like the 3Blue1Brown vibe but lighter.

Then:

Linear Regression
Multiple Regression

Slow foundational stuff, but it builds the base I need for later.

Today Specifically

Today was… structured in its own way.

I did my workout after a long time.
Bench press: 12.5 kg each side. Beginner numbers, but real movement. The point isn’t to be impressive; it’s to not let the body decay while the brain works.

I also made a GitHub commit and marked it as Day 1 of the streak.
Whether this streak survives or not is something future-me will reveal.

The Dream I Didn't Ask For

Last night, I had a dream where I was screaming into darkness.
Not metaphorically—literally screaming.

It was one of those dreams that leaves your body tired when you wake up, like your mind fought something your muscles couldn’t see.
Waking up after that wasn’t peaceful; it was more like resurfacing from a depth you didn’t plan to swim in.

I’m trying to get my mental health back on track, but moments like that remind me how volatile things still are.
Even so, I got up and worked. That’s something.

The Approach Changes Tomorrow

Starting tomorrow, I’m not doing four-day recaps or summaries.
Only what I did that day, clean and focused.

One day at a time.
No backlog. No compilation.

Where I Stand After Four Days

A solid chunk of DSA done
Math foundations moving steadily
Regression concepts revisited
GitHub streak restarted
Workout routine restarted
Mental state messy but functional
Still showing up

Day 3: Reflecting and pushing.

Blink — Mon, 08 Dec 2025 07:08:33 +0000

The first two days were all about laying foundations: 2 videos on vectors from 3Blue1Brown, and tackling Leetcode problems 217, 242, 1, 347, 128, 74, 704, and 153. Honestly? I didn’t approach them with full integrity. I tried, got stuck, and ended up looking at hints or solutions. My plan is to revisit each of these problems later with zero shortcuts—this is my metric for real progress. Linear and multiple regression also got some attention, but that’s just the technical side.

Beyond work, last night reminded me how fragile control can feel. I hit a near-breaking point but somehow held it together. I’m attempting a new perspective on life: not to succeed flawlessly, but to keep moving, keep observing, and keep recalibrating even if I stumble. The real experiment starts tomorrow, both with code and with myself.

Day 2: Low Energy, Work Done Anyway

Blink — Sun, 07 Dec 2025 09:19:54 +0000

Day 2 wasn’t impressive. I woke up feeling slow, and being at my grandma’s house doesn’t help. It’s one of those environments where everyone cares a little too loudly, and it kills any sense of pace. Not in a toxic way, just not compatible with how I function.

Still, I worked.

I watched the 3Blue1Brown video on vectors, which cleared things up more than any textbook explanation has. After that, I moved into hashing and stacks. Hashing still feels rough—it doesn’t click as naturally as other topics. No point pretending otherwise.

I solved LeetCode 217, 242, and 1. Then I had to take a nap because someone managed to irritate me enough to shut my brain down for an hour. After that, I went through linear regression.

That’s the whole list for today. Nothing dramatic. Nothing aesthetic. Just work done in a slow, unfriendly headspace.

I’m not tired, but I do feel a bit dead inside—more like I’m back in the same isolation I always return to. It’s not new, not scary, just familiar. Day 2 is over. Day 3 will be cleaner.

Day 1: Found the map

Blink — Sat, 06 Dec 2025 16:25:16 +0000

Starting from a clean slate feels strange, but necessary. After wiping everything and restarting the account, today became the first day of the new cycle. No legacy posts, no old signals bleeding into what I’m building now.

I started simple: posted the comeback, then sat down and built out my ML and DSA roadmaps for the month. Nothing fancy—just a clear structure of what needs to happen daily and what’s non-negotiable. I’m keeping both roadmaps practical and heavy on consistency rather than pushing unrealistic volume.

If anyone wants those roadmaps, reach out. You can mail me at somaykaush@gmail.com. I’m keeping them open because someone else might benefit from a clean, no-nonsense path.

There’s another piece to this: I’m not planning to rot in front of a laptop all month. Workouts start again on Monday. Coding + training—not for aesthetics, not for internet points, but because the mix keeps me stable. If you’re following along, you’ll have to handle both. It’s part of the rhythm I’m setting for this reset.

Day 1 wasn’t dramatic. It wasn’t loud. But it was a proper start.
The rest builds from here.

I archived everything

Blink — Sat, 06 Dec 2025 08:51:59 +0000

I archived everything.

Not because the old posts were bad, but because they belonged to an older version of me. A quiet reset felt necessary—strip the history, keep the intent, rebuild the signal.

So here’s the plan going forward:

Daily progress logs — what I’m learning, breaking, fixing.
Short builds — small projects that actually ship.
Long-term structure — OSSO and everything around it.
No polished guru energy — just the process as it happens.

I’m 19, still figuring it out, still switching lanes between CS, business, systems thinking. The internet moves fast, but documenting the climb keeps me accountable.

Fresh start. New cycle.
Let’s see what grows from this blank page.