Gaurav Sehrawat

This is why vintage engineering still scares modern tech

People are the big assets in this AI world, company's low performance/low performance is because of past choices and firing people won't fix it.

The Maps driving experience is also evolving with Immersive Navigation, featuring clearer visuals and intuitive guidance. You’ll be able to see the buildings, overpasses and terrain around you in a vivid 3D view, made possible with help from Gemini models. You’ll also be able to: 👀 See more of your route to prepare for what’s next. 🤔 Understand tradeoffs for alternate routes to pick what works best for you. 🛣️ Arrive easily with helpful details like parking and entrance information. Immersive Navigation starts rolling out today across the U.S. and will expand in coming months to eligible iOS and Android devices, CarPlay, Android Auto and cars with Google built-in.

I come back to this video every once in a while...

We just demoed fully working pull payments on @SUPRA_Labs . One-time. Subscriptions. On-chain. No UX hacks. Here’s what we built → Users connect once, set a mandate → Merchants define charge intervals daily, weekly, monthly, custom → Users see every active mandate in one dashboard → Cancel anytime. No contract. No custody risk. Web3 has had push payments forever. Pull payments, which power every SaaS, every subscription, every recurring bill, have been basically unsolved. Until now. We’re already integrating with 3 projects inside the Supra ecosystem. EVM, Solana, and multi-chain support is coming. We’ve abstracted away wallet connection friction, gas complexity, and the web3 UX nightmare, now merchants just ship, and users just pay. This is what Stripe + Visa/Master infrastructure looks like for stablecoins. Kudos to the team @RibbitWallet @nkaushal02 @Chatterjee_arn @RidhiOnChain @AdityaJyoti02

Today is my ONE month anniversary of me being on break, tinkering and being with family, friends and with myself. I had written an exit post, but then I converted that to blog, edited it, and finally published it here. Here is what it feels like. ajeygore.in/content/the-sp…

Every 6 months there is some major AI disruption! More to come infinitely.

@amitiitbhu Interested! And looking forward.

That tweet on RAG, MCP, LLMs, Agents, Fine-tuning & Quantization got way more attention than I expected. I want to teach these topics live, in an interactive way. • Completely FREE • Who this is for: Working professionals in tech • Prerequisites: None For folks who want to learn: Reply with "Interested" I'll plan to do it in batches, and share the details. Believe me, you'll learn it in the simplest way. And yes, I genuinely love teaching.

The game is changing. It's always been changing. Sometimes a little, sometimes a lot. But however the new board is set, you'll win nothing by turning bitter about whatever old advantages you've lost.

Software development is undergoing a renaissance in front of our eyes. If you haven't used the tools recently, you likely are underestimating what you're missing. Since December, there's been a step function improvement in what tools like Codex can do. Some great engineers at OpenAI yesterday told me that their job has fundamentally changed since December. Prior to then, they could use Codex for unit tests; now it writes essentially all the code and does a great deal of their operations and debugging. Not everyone has yet made that leap, but it's usually because of factors besides the capability of the model. Every company faces the same opportunity now, and navigating it well — just like with cloud computing or the Internet — requires careful thought. This post shares how OpenAI is currently approaching retooling our teams towards agentic software development. We're still learning and iterating, but here's how we're thinking about it right now: As a first step, by March 31st, we're aiming that: (1) For any technical task, the tool of first resort for humans is interacting with an agent rather than using an editor or terminal. (2) The default way humans utilize agents is explicitly evaluated as safe, but also productive enough that most workflows do not need additional permissions. In order to get there, here's what we recommended to the team a few weeks ago: 1. Take the time to try out the tools. The tools do sell themselves — many people have had amazing experiences with 5.2 in Codex, after having churned from codex web a few months ago. But many people are also so busy they haven't had a chance to try Codex yet or got stuck thinking "is there any way it could do X" rather than just trying. - Designate an "agents captain" for your team — the primary person responsible for thinking about how agents can be brought into the teams' workflow. - Share experiences or questions in a few designated internal channels - Take a day for a company-wide Codex hackathon 2. Create skills and AGENTS[.md]. - Create and maintain an AGENTS[.md] for any project you work on; update the AGENTS[.md] whenever the agent does something wrong or struggles with a task. - Write skills for anything that you get Codex to do, and commit it to the skills directory in a shared repository 3. Inventory and make accessible any internal tools. - Maintain a list of tools that your team relies on, and make sure someone takes point on making it agent-accessible (such as via a CLI or MCP server). 4. Structure codebases to be agent-first. With the models changing so fast, this is still somewhat untrodden ground, and will require some exploration. - Write tests which are quick to run, and create high-quality interfaces between components. 5. Say no to slop. Managing AI generated code at scale is an emerging problem, and will require new processes and conventions to keep code quality high - Ensure that some human is accountable for any code that gets merged. As a code reviewer, maintain at least the same bar as you would for human-written code, and make sure the author understands what they're submitting. 6. Work on basic infra. There's a lot of room for everyone to build basic infrastructure, which can be guided by internal user feedback. The core tools are getting a lot better and more usable, but there's a lot of infrastructure that currently go around the tools, such as observability, tracking not just the committed code but the agent trajectories that led to them, and central management of the tools that agents are able to use. Overall, adopting tools like Codex is not just a technical but also a deep cultural change, with a lot of downstream implications to figure out. We encourage every manager to drive this with their team, and to think through other action items — for example, per item 5 above, what else can prevent a lot of "functionally-correct but poorly-maintainable code" from creeping into codebases.

10-minute delivery in India isn't magic. It's really good engineering. @albinder and @letsblinkit built tech that most people never see. We went deep on how it works.

Always give credit where it’s due. Call out people who have helped you even tiny bit, that goes long way. Expressing gratitude is the most valuable thing you can do for yourself.

As a broker, there are rare days when risk management simply doesn't work, when markets move so violently that traders lose more than their entire initial margin. When this happens, both the trader and the broker are sitting ducks with no way out. Yesterday was one of those days in commodity markets. All major metals hit lower circuits—the maximum they can move in a day. Silver crashed 30%, Gold 15%, and others followed. Btw, Natural gas was on an upper circuit. In our 16 years of operations, we've only seen something like this once before: when Crude oil closed at a negative price during COVID. But that was just one commodity, and commodity trading wasn't nearly as popular as it is today. What happened in commodities yesterday can happen in equities too; we saw it in 2008. The lesson is simple but critical: only trade with money you can afford to lose. You can trade successfully for a decade and lose it all in a single day if you're not properly managing risk. There's no margin call, no exit opportunity when markets gap through circuits like this.

I come back to this Ed Sheeran video once in a while...

Absolute power corrupts people!!

VC vs. founder

.@TheOceanCleanup is now intercepting 2 to 5% of global plastic pollution. I think we should be able to double again this year.

Defending my Spring Boot Java app that uses 64GB RAM to return { "status": "ok" }

【Horror】A Collection of Accidents That Actually Happened in Reality

i never want to read any other way again.