> Got denied leaves for offcampus by college TPM because the same company came for oncampus as well, and my TPM got to know it. > Was told basically "CGPA dekhi hai teri?" > surprised_pikachu.jpeg > Got shadowbanned in placements > Got a 10 LPA job from oncampus because recruiters got to know me from a hackathon > Left offer, one of my friends got in my stead, me happy > Got $100k USD WFH offer, left that to join another promising startup (WFH) > Punished by being forced to stay an extra sem > TPM made an example out of me by sending a mail (can only DM SS for obvious reasons) > Became part of my batch lore, instant popularity - asked out even on LinkedIn lmao (but I already won in life with my gf) > dab_while_crying_internally.jpeg > handled full time job with extra penalty courses and B Tech Final Year Project > Worked on Data conflicts and hallucinations in LLMs as my BTech thesis > Started building @getalchemyst from college room, with @uttaranxnayak > Resigned from aforementioned startup job > Advisor fac didn't know what I was working on, gave me less marks > Left my job, raised preseed funds > Btech thesis got published in Springer > happy_happy.jpeg > Fac indirectly apologized later on > never_give_up.jpeg

We're taught to keep our "minds open" and "mouth shut" as a fundraising exercise. Especially us younglings who have little to no signals of credentialism.

I so wholeheartedly second this. A small rant. <rant> Last week we onboarded an enterprise customer - an Indian F500. Two things as usual: - Delay in wiring money - Paid for the product, demanded extra help without money. When we allowed it, started to scope-creep. 3 minutes into the meeting: > One of the CxOs: This is already done by this XYZ company (name hidden for obvious reasons). What are you providing extra to me? > Me: Explains them, shares the exact document built for questions like this. > Him: Yeah, but why would I trust you guys over them if you don't deliver the outcome for me? > Me: You paid for the product right, that too only after sufficient testing. > Him: That's not a professional way to talk, we're implementing software on our end for more than 25 years... > Me: Excuse me? I'll be gladly helping- provided you pay for the engineering hours that we'll need to calculate. > Him: See, I see a lot of US companies also doing the same - why won't I procure them? >Me (really flabbergasted): Because they won't change US Data residency to India, and you need strict data compliance within a budget that won't fit the margin of errors. > Me (seeing his silence): For more technical reasons, we're better than them for your use case, as your product team also pointed out before. > Him: Let's do one month of free pilot first, US companies like SalesForce and Microsoft are more reliable who've been doing this. > Me: First pay up please. We've always regarded dev platforms are the tyres on which the cars of outcome-based agents run. Nowadays we've started to joke internally: "Inn logo ko kya tyre bechenge?" This is the current state of mentality across Indian companies. The pimps at IT services companies have made a joke of builders by spreading this virus of doing mediocre work at the least prices. DO NOT CALL THIS SLAVERY EMPLOYMENT. AND THE SO CALLED "VISIONARY VETERANS" ARE LITTLE MORE THAN BROKERS WHO LIVE OFF LABOUR ARBITRAGE. </rant>

indian sovereign tech enthusiast brain: enthusiast - “kotak bhai, tum innovation karo” kotak - “lekin mai toh bank chalata hu” enthusiast - “desh bhakti naam ki cheez hai ki nahi tumhare andar buddhe, chalo fighter jet banao kal se”

I am amused that many in India who call themselves “nationalists,” along with those accused of not being nationalist enough, are now pushing for strong government intervention through expanded IndiaAI Mission-style Sovereign AI initiatives or new national tech missions in response to restrictions like those on Anthropic models. To me, this mirrors the License Raj socialism of the Nehru era, which almost everyone, irrespective of political affiliation, criticises as a bloated, corrupt, and inefficient system that strangled enterprise and innovation. The difference is that under Nehru we had public sector giants supposedly built for national development. The new political mantra seems to be to create state-funded and state-supported private sector giants owned by cronies close to government power. Even when the government is not directly funding them, the expectation is that it will backstop them with cheap credit, subsidies, and favourable policies. This approach is now being rebranded as the real “Swadeshi” model. In practice, it amounts to crony capitalism on steroids. Neither the “nationalists” nor their critics, often branded as “globalists” or worse, “anti-nationals,” seem interested in the solution that would truly benefit India and its diverse people, who are desperate for honest opportunities to use their talent, work hard, and succeed. The real moral solution for India is an ultra-minimal government combined with a maximally open and free market, operating under a system of law and justice that everyone can trust. Such a libertarian-style system would remove the bureaucratic red tape and excessive controls that currently stifle innovators and kill the tinkering and experimentation culture essential for genuine progress. Such a system would also unleash India’s hidden and suppressed entrepreneurial energy from the bottom up. In my opinion, true nationalism and true Swadeshi mean removing the barriers that stop grassroots Indians from building and competing freely, rather than propping up big-government central missions and controls that favour only a few cronies.

I've been saying this time and again - companies like us are starved of compute and funds. Hand compute grants - and see magic happen.

We went that route, but fundraising is hard. Especially with folks who don't have connections in the market. India isn't ready to widely fund SV level innovation yet.

RT @HarveenChadha: I wish I could do a podcast with people who were against building frontier models on why they didn't foresee this coming…

anuran.dev/projects/langtach… For most agent-building usecases, using Langchain (especially in TypeScript) is an overkill. Why not use something lightweight that's also composable like LCEL, while cutting down all abstractions? Made a package, did a quick writeup - check this out once!

Speedrun started. Some cool stuff incoming.

And now we're (almost) at 6k. HuggingFace downloads are by no means a performance metric, but shows that we're going in the right direction ^_^ Will package this into something VERY useful soon!

We got selected for Razorpay's The Next Big Thing Cohort 🚀 Alchemyst AI is building the context and memory backbone for enterprise AI agents. We help AI agents remember, learn, and operate with the context businesses actually need. Built by two 20-something founders from India🇮🇳 Already working with ambitious teams pushing AI into production. Just getting started GUYS 🚀

How aggressively competitive are you? NVIDIA: We made an entire architecture that can be switched on/off at the logical level depending on how an if-then-else evaluates.

Now that I've got to know about this cool thing called Coccinelle, I'm really intrigued by this very niche field that promises to port entire codebases from a language to another *DETERMINISTICALLY*. There's also a ton of fascinating history behind this. Would be diving deep into this next weekend!

Data engineers know data drift - Zillow was the prime example of what happens when data drift gets out of control. Now, what happens when semantic drift occurs for and with your AI agents? The same rules apply, but how deep does it go? Here's a quick weekend read for you. Would love everybody's feedback on this! anuran.dev/blog/semantic-con…

RT @Hesamation: Mythos, regulate something. make no mistakes.

Feeling humbled after getting loads of reality checks. SO MUCH TO LEARN, SO MUCH TO DO.

Well, search mine.