Joined January 2023
- Tweets 371
- Following 70
- Followers 5
- Likes 178
204 Photos and videos
Jun 13
Your p50 dashboard is lying about your agents.
A 20-step agent hits tail latency ~18% of the time. Not bad code. Just math: ~1% odds per call, compounded over 20 steps.
The median never sees it. Users do.
#SoftwareArchitecture #DistributedSystems #SRE #Reliability
8
Anil Kurmi retweeted
Jun 8
AI won't take your job. It's quietly changing what your work is worth.
I spent months on how people actually survive automation waves. Wrote it up as a free 6-part series — no paywall. Thread of all six 👇
#AI #FutureOfWork #CareerGrowth #Upskilling
1
1
2
22
Jun 8
AI won't take your job. It's quietly changing what your work is worth.
I spent months on how people actually survive automation waves. Wrote it up as a free 6-part series — no paywall. Thread of all six 👇
#AI #FutureOfWork #CareerGrowth #Upskilling
1
1
2
22
Jun 8
Part 3 — The Structural Shift. What the labor data really says:
practiceoverflow.substack.co…
Part 4 — Global & Industry Patterns. The survival map across 14 sectors:
practiceoverflow.substack.co…
1
7
Jun 8
Part 5 — The Builder's Playbook. 30/90/365-day moves for engineers, freelancers & IT pros:
practiceoverflow.substack.co…
Part 6 — The Knowledge Worker's Playbook:
practiceoverflow.substack.co…
Which of your tasks are gaining value as AI improves — and which are quietly losing it?
4
Jun 7
"Will AI take my job?" is the wrong question.
Ask: which of my tasks got cheaper, which got more valuable?
AI reprices tasks inside a job, not the whole job. Move toward work AI can't sign off on 👇
practiceoverflow.substack.co…
#AI #FutureOfWork #AICareers #TechCareers
4
Jun 6
AI job loss is not one story.
Europe: hiring cuts.
China: robot supply chain.
India: mid-tier IT beats scale.
Philippines: BPO pivots to capability.
Same AI. Different labor markets.
The moat is capability per person.
#AI #FutureOfWork #Automation #IndiaIT #TechCareers
9
Jun 6
The best systems don't hide risk.
They move it where teams can control it.
This week: containment over fake fixes, graceful overload over heroics, agent runtimes over tool sprawl, living runbooks over PDFs.
#SoftwareArchitecture #DistributedSystems #SRE #AIInfra
5
Jun 6
AI Weekly W23: Microsoft shipped 7 in-house MAI models. Uber capped AI coding at $1,500/mo. Ladybird closed PRs. Hackers asked Meta AI for Instagram accounts and got them.
Capability is racing. Trust, budget, and auth aren't.
#AIWeekly #AI #Microsoft
6
Jun 2
Job-scheduler LLD: 40 min burned on Redis vs SQS. Real rubric: what happens when a worker dies mid-task? Lease heartbeat idempotent retry DLQ = yes. Miss it, stay senior. What failure mode did YOUR interviewer drill? #SystemDesign #InterviewPrep #BackendDev #Career
11
Jun 1
Killed our 200-row golden set for property-based evals on tool-call JSON. Hypothesis fuzzed 5K variants, caught 3 schema drifts the GPT-4o→GPT-5 snapshot tests missed. $0.40/run. Anyone else ditching golden sets for fuzzing? #LLMOps #Evals #AI #SoftwareEngineering
2
May 31
Cross-region sync replication: 280ms per write. Switched to LOCAL_QUORUM async cross-DC with per-region idempotency keys — same correctness, 18ms p99. Strong-everywhere is almost never what you actually need. #DistributedSystems #SystemDesign #BackendDev #Cassandra
5
May 30
18KB tool response, 80% crammed back into the next prompt — our LangGraph agent kept blowing context. Wrapped every tool output in a Pydantic schema. 18KB→700 bytes. Task completion 64%→91%. #LangGraph #LLMOps #AI #Engineering
1
10
May 26
Armin Ronacher begged OSS filers to stop piping bug reports through an LLM. Fake repros. Confident-wrong root causes. Error lists that might not matter. AI didn't ship a feature for maintainers — it shipped a queue of plausible garbage. #AI #OpenSource #LLMs #DevEx
5