JUST 365 DAYS AGO: I was : 👉🏻Underweight(44kgs) 👉🏻Under confident 👉🏻no social skills 👉🏻Addicted 👉🏻not future oriented Now: 👉🏻Somewhat healthy(60kgs) 👉🏻Can talk to anyone 👉🏻No addiction 👉🏻Focusing on myself and my career 👉🏻Building meaningful connections You just need a WHY!

Late night building matterzero : legaltech saas for immigration teams 🥳

I would trust someone who have vibe coded apps deployed live than someone with leetcode streaks 🙂‍↔️

Gridline 🌬️ Few big problems with this idea and research i can say is : 1. Getting messy files into w consistent format 2. Matching the right forecast to the right actual 3. Explaining bad or missing data 4. Making AI careful and evidence based 5. Protecting customer data

Anthropic doing everything for the valuation 😭 Prediction: OpenAI will possess more danger to humanity in future than Anthropic, no AI should be slowing down now it already destroyed millions of life so no point now stopping until we see the ultimate of AI !

I love AI not because it makes coding easy , I love it because it helped me explore industries i never thought i can learn and build upon 🧋 Currently i am building GridLine basically a high density intelligence and telemetry platform for energy market participants 🌬️🎐 no guesswork! AI indeed gave a super power to everybody to be curious and build for an Industry not random movie app <3 Use it! Build it!

LegalTech is amazing! Scaling context across hierarchical legal docs The signal lives between documents, not inside them. Naive chunking or long-context dumps both break this at scale. What fails • Flat chunks sever parent-child and cross-doc links • Long context degrades attention and gets expensive • Retrieval returns isolated snippets with no version or amendment awareness How we scale it 1. Document graph first — Parse every matter into nodes (docs) + edges (amends, schedules, supersedes). Track version lineage and effective dates. This becomes the source of truth. 2. Hierarchical multi-hop retrieval — Chunk at clause → section → document level. Retrieve leaves, then always expand to parents + linked docs. Rerank over the neighborhood, not raw top-k. 3. Structured extraction layer — On ingest, pull defined terms, governing clauses, and change events into a structured store. Hybrid retrieval: dense search + precise lookup. 4. Matter-level agent : Agent walks the graph in short, tool-grounded steps instead of one giant prompt. Example Query: “What’s the current payment term under the master agreement?” • Graph finds Master Agreement v3 + Amendment 2 (supersedes v2) + Schedule B • Pulls only the payment clause from each + defined terms • Agent resolves: Amendment 2 controls → net 45 → references Schedule B for early payment discount No full document dumps. Just the relevant lineage.

I was trying to do a bit of research on whats happening around the world before vs after the AI revolution. Its kinda tough for me to understand on a broader view to all the founders building around AI. I want to know what is the most common bottleneck? Apart from Distribution

I have 4 years of experience in engineering but swear to god i never reversed a linkedlist 🙂‍↔️ Its great to see Hackerrank is changing too, leetcode and DSA is trash!

Yesterday I Implemented idempotent ingestion tasks with retry and state recovery handling for interrupted worker jobs!! The new version of matterzero would be a killer 🔥 the legaltech space have so many things missing and so many things i personally hate it all needs a revamp..

Been off due to puja and got bad cold now back at grinding! Meanwhile gave some interviews with some cool founders loving the energy of where the industry is moving. Lets gooo 🔥

Yesss!

Down the line in future, i want to own a video games company or work for a video game company,not because its cool or anything but amount of fun it can add to humans life apart from all this weirdest things happening in the world. I would be super fulfilled as a human. No B2B SAAS can ever do what a video game can do

To all the vibe coders, If you all having a long running background process for a task and there are several tasks waiting in the queue, please add a checkpoint after each successful task :) big learning! Also please add a heartbeat on the process that can literally save you thousands of dollar 💵 And if there is retries in AI processing add checkpoint and save and retry should happen from that saved point not whole process again.. By doing this i saved soo much money to another vibe coded platform lately!

I watched “The playlist” on Netflix last sunday and since then i got addicted to system architecture i literally now on my free time go and ready Engineering blogs. My recent favourites are : 👉🏻Legora’s Engineering blog 👉🏻Harvey’s Engineering blog 👉🏻Posthog’s Engineering blog (mainly the “How we made feature flags faster and more reliable”) Crazy time to be alive i mean just think and solve the problem for implementation and help you think with more clarity we have our claude and codex 🙈

While solving any tech problem with AI agents just follow a simple rule to avoid AI slops (I hate AI slops) Start with : Goal -> Gap -> Scope -> Constraints Thank me later, please dont use any free skills available for this! Reason is rvery product have a different outcome it should be related to your own idea but ofc you can create a skill or .md file for this specific task. Dont create slops

Shit man woke up with cold and slight fever 🤒 i have a big meeting coming up this friday kinda nervous also gotta prepare!

Confused , what should be the extension i get? 👉🏻 .si 👉🏻 .com 👉🏻 .in not .ai cause too expensive lol!