Abhay

894 posts

Abhay banner
Abhay

Abhay

@bothra90

I am interested in Stream Processing, Distributed Systems and Databases. Co-founded @FennelAI. Currently @Databricks

San Francisco Bay Area Katılım Ağustos 2009
2.1K Takip Edilen576 Takipçiler
Sabitlenmiş Tweet
Abhay
Abhay@bothra90·
The next time you’re prioritizing, create estimates first! Let me explain why: (1/n)
English
2
2
18
0
Sam Lambert
Sam Lambert@samlambert·
If we held an event where we dug deep into PlanetScale's architecture to show you why we are so resilient would you attend?
English
140
9
898
30K
Diptanu Choudhury
Diptanu Choudhury@diptanu·
I get a lot of questions about why Kubernetes isn’t the right foundation for sandbox infrastructure. It was built for stateless micro-services with predictable traffic patterns, and databases that run forever. Sandboxes are different, they’re high-throughput like batch jobs and stateful like databases. You don’t want your vibe-coding agent to install all your app’s dependencies, download code, etc. every time it starts. You want the sandbox, with the agent in it, to start immediately, with code, dependencies, agent memory preserved from the last session. Doing this with Kubernetes at scale is hard, and even with a lot of hacks it would be working around warm pools.
English
10
4
49
6.4K
Diptanu Choudhury
Diptanu Choudhury@diptanu·
@kellabyte Well lol I didn't want to dump on etcd. Kids these days don't know what etcd is Kelly :D
English
1
0
4
338
Acquired Podcast
Acquired Podcast@AcquiredFM·
@igordownunder There's a good chance that this company is incredibly important to you personally, but you rarely think about it. The founder has proven that less does indeed lead to more. (Dropping Monday)
English
9
3
84
27.6K
Igor
Igor@igordownunder·
@AcquiredFM any hint on new episode ?
English
1
0
1
3K
Abhay retweetledi
Branko
Branko@brankopetric00·
Kubernetes Secrets are not encrypted. They are base64 encoded. We have been calling them secrets for ten years and the auditors have not caught on.
English
31
42
747
65.7K
Abhay retweetledi
Billy Binion
Billy Binion@billybinion·
Bill Gates helped make personal computing mainstream. Sergey Brin dramatically expanded access to information. Jeff Bezos gave the world quick & cheap access to goods. AOC gets a taxpayer-funded salary to do political theater. Yes, it is possible to earn a billion dollars.
Marco Foster@MarcoFoster_

AOC: “There’s a certain level of wealth and accumulation that is unearned. You can’t earn a billion dollars. You just can’t earn that. You can get market power, you can break rules, you can abuse labor laws, you can pay people less than what they’re worth, but you can’t earn that”

English
149
399
3.5K
111.4K
Abhay
Abhay@bothra90·
@rakyll I love Go, but I think the problem is that Go makes concurrency too easy to use as a tool
English
0
1
1
695
Abhay retweetledi
Mario Zechner
Mario Zechner@badlogicgames·
People of pi. I'm removing Gemini CLI and Antigravity logins from pi. Welcome to 2026, the year of the end of subsidies.
English
74
30
1.4K
139.3K
Jaana Dogan ヤナ ドガン
Go provides lightweight threads and a variety of synchronization capabilities as core language features for convenience and a unified library space. Go never claimed that concurrency is easy.
James Ward@JamesWard

Generally developers think of Go as being great for concurrency. Its not. JVM approaches are vastly superior. And even some of the best in the whole industry when you include virtual threads, structured concurrency & Effects.

English
15
21
341
41K
Abhay retweetledi
Fatih Arslan
Fatih Arslan@fatih·
Kubernetes isn't a scam. People don't just realize what it is, and what it brought to the world. What Kubernetes did to the world is to teach and bring Control System theory to the masses. With control system, you could run software at scale like never before. If you design your software in closed feedback loops, you can have, just like a machine, an ongoing stable system running for 7/24, that can self recover and steer itself. People trying to use other orchestration systems, had to work and implement all of it themselves. And most of them didn't had proper primitives, so it was very brittle. With Kubernetes, you have /status, the reconciler/controller-runtime framework, requeues and CRD's. If you use all of these together, you can build a feedback loop, and apply control systems knowledge. And with Google's push, it became the winner. There is a really nice book about it: "Designing Distributed Control Systems: A Pattern Language Approach". It's actually about machines, not software (like how to build proper big machines that can run 7/24). But if you read it, you immediately see how the patterns in the book described, are actually primitives used by Kubernetes ecosystem.
Zack Kanter@zackkanter

Kubernetes

English
60
139
1.9K
218.2K
Abhay
Abhay@bothra90·
@nipun_s6t @vishalmisra I’ve switched almost entirely to Codex, who I also find to be more intelligent
English
1
0
0
54
Nipun Sehrawat
Nipun Sehrawat@nipun_s6t·
@vishalmisra That's my biggest sticking point. Yesterday I wanted CC to work non stop for all 6 hrs I was out hiking Muir woods. But it called quits after an hour.
English
1
0
0
43
Vishal Misra
Vishal Misra@vishalmisra·
I have noticed this as well in the past few weeks - Claude Code is very keen to “call it a day and resume tomorrow?”. I think this is deliberate post training - there is a lot of token/cost pressure on Anthropic is what I surmised.
Nate Silver@NateSilver538

This is interesting. Claude keeps telling me to "wrap it up" when there are unresolved problems in the model I'm building. It probed it, and it says its training is flawed for this sort of use case, and I should flag to @AnthropicAI.

English
6
0
11
1.8K
Abhay retweetledi
Mitchell Hashimoto
Mitchell Hashimoto@mitchellh·
Now I understand the full picture. The cleanest fix is... But actually, the real fix simpler... Actually wait. The best fix: Now the real fix. Actually, let me reconsider. OK Key finding: Wait I need a hardware device I can physically punch to stop the agentic session.
English
179
160
4K
115K
Abhay
Abhay@bothra90·
This resonates with what and many others have been feeling. Codex is a very high IQ colleague where as Opus tries hard to seem intelligent, but given hard problems goes in circles or simply gives up.
antirez@antirez

The difference between the two, for serious engineering work, is simply brutal. Claude Code with Opus is, when the task at hand is very complicated, borderline useless, while GPT 5.4 can do a reverse engineering mixing: hardware knowledges, major disassembly skills, and so on.

English
0
0
0
127
Abhay
Abhay@bothra90·
@clairevo What’s your take on Alpha schools?
English
0
0
0
14
claire vo 🖤
claire vo 🖤@clairevo·
I have experienced private, parochial, and public for my kids. 90% overlap on academics. Public is 10x more resourced. Parochial best sports. Private wins on PX (parental experience.) Great teachers in all, meh teachers in all. Know what you’re paying for if you go private! It (in my limited experience) is not differentiated academic outcomes.
Nick Maggiulli@dollarsanddata

Private schools are the most expensive placebo in America. Nowhere else will you pay $250k+ for something that has so little impact on school achievement. My latest on why private school isn't worth the cost: ofdollarsanddata.com/why-private-sc…

English
16
1
96
32.6K
Abhay
Abhay@bothra90·
@threepointone Without a human/agent in the loop, doesn't codemode enable one MCP tool injecting malicious prompts into the next one? I am thinking something like SQL injection attacks.
English
1
0
1
75
sunil pai
sunil pai@threepointone·
code mode: let the code do the talking (aka, after w/i/m/p) wherein I ponder the implications of every user having a little coding buddy, and every "app" being directly programmable on demand. sunilpai.dev/posts/after-wi… lmk what you think.
sunil pai tweet media
English
43
26
284
141K
Abhay retweetledi
Ryan Petersen
Ryan Petersen@typesfast·
Everyone is talking about the impact of closing the Strait of Hormuz on oil and gas. I haven't seen much discussion of the war's impact on helium. Most people think of helium as being a fun gas you use for children's balloons. But helium is actually one of the most important industrial gases, used in rockets, MRIs, quantum computers, and most importantly, in the production of semiconductors. Qatar produces ~30% of the world's helium as a byproduct of its natural gas wells, and that supply is now cut off from the global market. 🧵
English
144
781
4.3K
691.1K
Abhay retweetledi
dax
dax@thdxr·
sent this to the team today everything great comes from being able to delay gratification for as long as possible and it feels like we're collectively losing our ability to do that
dax tweet media
English
256
704
6.9K
980K
Abhay
Abhay@bothra90·
Feature stores have a new name! “Analytical Context Databases” doesn’t roll off the tongue though.
Tomasz Tunguz@ttunguz

Enterprises learned a lesson from cloud data warehouses. They handed over both data & compute, then watched as the most strategic asset in their business, how they operate, became someone else’s leverage, which created an opportunity for Iceberg. Fool me once… Leaders have recognized their companies need a new system of record for AI agents in the form of a context database. There are two different kinds of these context databases : (first image) Operational context databases store standard operating procedures & institutional knowledge : when a customer calls about resetting a password, when legal reviews an NDA with a new prospect, when HR answers questions about options vesting for a new hire. All of these processes represent trade secrets & intellectual property, which are key assets for a business. Capturing them from employees ensures continuity in processes & builds a sustainable, defensible asset. Analytical context databases are a semantic evolution of semantic layers : they contain definitions & calculations for metrics like revenue or customer acquisition cost. Semantic layers told AI what data meant. Analytical context databases teach AI how to reason about it. Steven Talbot’s recent piece on Omni’s agentic analytics architecture describes : a coordinator mechanism, which decides which tool to use next based on the question, the results, & what’s already been tried. The key to both operational & analytical context databases isn’t the databases themselves. It’s the feedback loops within them. Steven’s system adapts mid-flight, retries when things break, or stops when it has something useful to show. This creates an ever-improving cycle of accuracy. Accuracy creates trust. Trust creates adoption. Adoption creates more feedback. Companies that develop the best feedback loops will build the most valuable context databases. Context databases enable the future of process automation, representing the real promise of AI within the workforce. It’s the evolution of RPA (robotic process automation), but it’s RPA & process discovery injected with non-determinism. This non-determinism is essential for the success of AI agents. It allows for exception handling, forestalling one of the failure modes of the first generation of RPA. AI agents are excellent at ingesting large volumes of content & reasoning about them. The move from manual context engineering to automated context platforms is inevitable. Context databases will be sold as standalone products & bundled. Enterprises will come out of this transformation for the better : with evolving systems that improve over time. tomtunguz.com/operational-an…

English
0
0
1
209