Job post alert! π¨
Looking to hire a kickass UI/UX designer with great taste and abundant user empathy.
In person role in bangalore. 50-60 LPA, Ideally 5+ years of experience.
We are building a SOTA Accounting AI agent to replace all the manual grunt work that an Accountant joylessly does.
DM me or mail your portfolio to [email protected]
Mukesh Ambani literally bet Reliance on Jio in 2016
From a close to 0% revenue share then, Jio is now 40% of the company's revenue
$50B of audacious investment in disrupting telecom allowed Reliance to diversify away from energy
Underrated business owner
We experienced an outage at Coinbase last night, which is never acceptable. The root cause was a room overheating in an AWS datacenter when multiple chillers failed. We design our services to be redundant to downtime in any one AWS Availability Zone (AZ), and most of our systems worked this way last night, but not all.
Our centralized exchange did not. Exchanges have unique architectures that optimize for latency and co-location of clients. It is possible to make exchanges resistant to AZ failures, but this can introduce latency delays that are not desirable along with breaking customer co-location. Given this incident, we'll revisit these tradeoffs to ensure we're giving you the best possible venue to trade. At a minimum, the duration of an outage should be able to be reduced considerably when an AZ move is needed.
Thank you to the AWS and Coinbase teams for working through the night to mitigate the issue. Weβll share the detailed technical summary once it's ready.
Codex shares the diff and summary of it's cloud agent work in a publicly shareable URL directly into Slack.
Think it's powered by the same chatGPT share mechanism. Terminated Github access for Codex agent.
WTF @OpenAI .
For the first time, seeing Β§ symbol in the plan documents in claude code. Given that the symbol is mostly used in Legal documents, Feels like models are absorbing legal language patterns from their training data.
Would be interesting to see emerging capabilities because of more and more inter disciplinary knowledge in larger and larger models.
The most interesting breakthroughs tend to happen at the intersections.
We rolled out Aqqrue to beta users β itβs already handling accounting for ~50 businesses.
But building agentic systems = new problems.
More code β more edge cases.
Non-determinism β bugs hide behind βsuccessfulβ outcomes.
Silent tool failures
Inefficiencies
Weird execution paths
Gaps masked by results
Evals help, but I/O correctness isnβt enough.
So we did something different:
π Let the agent report its own issues.
Now it flags failures, inefficiencies, and suspicious behavior in prod + staging β sends directly to Slack.
Result:
We catch root causes in real time, before more users feel it.
If youβre building AI agents:
Add an internal debugging + issue-reporting layer.
Game changer.
To stop ants coming in to your house leave a saucer of milk outside. The adult ants drink it & it has an effect on ant reproduction. The young are born without toes so they can't climb in to your
cavity walls.
This effect is called lack toes in toddler ants.