Kushagra Mathur

Software Engineer | AI Builder
Turning ideas into products with code.
Sharing code, system design & career insights

[Kushagra Mathur @CoderGhost37]

@nikitabier The scary part isn't the voice, it's that banks still treat caller ID and "confirm your OTP" as identity. Calling back on a known number should be the default, not a tip buried in a PSA

[Kushagra Mathur @CoderGhost37]

@piyushgarg_dev The part people miss is that memory is really a retrieval problem.
Deciding what to save and when to inject it back into context matters more than the model, and bad memory is worse than none

[Kushagra Mathur @CoderGhost37]

@theo Multi-model review only works if one model's confident nonsense can't outvote the others.
Do you weight the reviewers or just flag where they disagree?

[Kushagra Mathur @CoderGhost37]

@striver_79 The rewrite always looks clean because it hasn't met production yet.
Those ugly if-statements are years of bug reports nobody wrote down

[Kushagra Mathur @CoderGhost37]

@striver_79 This is way closer to the actual job than LeetCode. Half the skill is knowing when the AI's fix is wrong, and you can't do that without knowing the algorithm

[Kushagra Mathur @CoderGhost37]

@Sumits_X I'd put idempotency and retries above microservices on this list.
A lot of prod incidents turn out to be a retry storm or a duplicate write, not a scaling problem

[Kushagra Mathur @CoderGhost37]

@rileybrown Every "is the new model dumber" thread is half real regression and half us getting used to it. A small personal eval set is the only way to trust the vibe

[Kushagra Mathur @CoderGhost37]

@theo Six months and it already feels like a different era. The hard part now isn't picking a model, it's keeping your evals and prompts from rotting every quarter

[Kushagra Mathur @CoderGhost37]

@shydev69 @poteto 200 PRs in a week is wild, but the real flex is the review loop behind it. How much is the skills vs you getting stricter about what "done" means?

[Kushagra Mathur @CoderGhost37]

@GergelyOrosz @Lethain The "why" only showed up once agents needed the context. A lot of tool migrations look pointless until the API becomes the real product.

[Kushagra Mathur @CoderGhost37]

@cneuralnetwork Cheap tokens change how you design, not just your bill. At that price you can run 3 attempts plus a verifier and still pay less than one frontier call.

[Kushagra Mathur @CoderGhost37]

@ThePrimeagen One test that runs every day beats fifty that run when someone remembers. The schedule is the real win, the coverage will follow.

[Kushagra Mathur @CoderGhost37]

@GergelyOrosz Shipping inside the app people already spend hours in is a bigger moat than a slightly better model. Being first in the EU on top of that is a sneaky good combo

[Kushagra Mathur @CoderGhost37]

@emollick Harder to inject in a single chat, sure. The real risk is agents reading untrusted pages and tool output while holding write access. Better models don't remove the need for permission boundaries

[Kushagra Mathur @CoderGhost37]

@GergelyOrosz The flip side is that the premium only goes to people who keep re-learning. Engineers who stopped updating in 2022 are already finding that out in this market.

[Kushagra Mathur @CoderGhost37]

@arpit_bhayani On most teams I've seen, coding was maybe 30% of the cycle. The rest is review, CI and waiting on decisions, so the real bottleneck has quietly moved to PR review.

[Kushagra Mathur @CoderGhost37]

@NikkiSiapno The part that changes most in agentic SDLC is the spec. A vague ticket used to cost a day; now it costs a confidently wrong PR, so writing clear specs is the new senior skill.

[Kushagra Mathur @CoderGhost37]

@levelsio Did this on my side project last month. Zero open ports apart from the tunnel, and the auth log went from thousands of bot hits a day to silence. Cheapest security upgrade I've made.

[Kushagra Mathur @CoderGhost37]

@huge_icons Decide what not to build. Claude Code will happily ship whatever you ask for. Knowing which ticket is a bad idea, and explaining why, is still the job.

[Kushagra Mathur @CoderGhost37]

@shadcn A foldable iPad Mini with a real terminal and SSH would replace my laptop for half of on-call. Nobody needs a third camera to review a PR.