ASU / SYSTEMS Available for hard problems

Case board 01 · systems / writing / life

Last mapped · 2026

00 / Profile01 / Systems + Work02 / Field Notes03 / Life

Subject / 01

Backend systems engineer

  • Java + concurrency
  • Production reliability
  • Agentic AI systems
Asu Singh
Asu / system owner

Senior Software Engineer

Working thesis

I make complex systems behave.

20×trade throughput
90%downtime removed

request → queue → worker → recovery

design for the failure path

p99< 120 ms

Observability / live

latency · errors · queue depth

draft → challenge → evidence

keep the useful footnotes

APIqueueworker Aworker Bworker CDLQretry ×3closedopenhalf-openprobe 1%

Field notes / Blog

Systems, AI & performanceLong-form writing is part of the map →

Off the clock / Life

Ideas, books & elsewherePersonal notes, kept human.

Real-time systems · recovery paths · measurable outcomes

Follow the thread

Case board / 00

Profile / System Owner

Engineer behind the board

A senior software engineer focused on making complex production systems understandable, recoverable, and fast.

Evidence / 01

Core craft

Backend systems

Java, concurrency, distributed workflows, and performance-sensitive services.

Evidence / 02

Operating principle

Trace the evidence

Start with observable behavior, follow the failure path, and measure the recovery.

Evidence / 03

Current direction

Agentic AI

Building useful agents with production-grade constraints, tooling, and guardrails.

Zoom out. Follow the next board →

Case board / 01

Systems / Work

Evidence from production

High-throughput systems, deliberate failure paths, and recovery that can be measured.

Evidence / 01

Low-latency trading

20× throughput

Concurrency, partitioning, and backpressure without losing ordering.

Evidence / 02

Reliability engineering

90% less downtime

Recovery paths, replayable queues, and safer operational defaults.

Evidence / 03

Agentic AI systems

Tools with guardrails

Observable agents designed to operate inside real production constraints.

Zoom out. Follow the next board →

Case board / 02

Field Notes / Blog

Writing from the failure path

Practical essays about systems, AI, performance, and the decisions hidden between the diagrams.

Evidence / 01

Failure paths first

Systems essay

Why the recovery story should be designed before the happy path.

Evidence / 02

Agents under load

AI engineering

What changes when an agent becomes part of a production control loop.

Evidence / 03

Concurrency field notes

Performance

Patterns, bottlenecks, and lessons gathered from Java systems.

Zoom out. Follow the next board →

Case board / 03

Off the Clock / Life

The human side of the map

Books, unfinished ideas, visual observations, and the side quests that make the systems work better.

Evidence / 01

Reading pile

Books + marginalia

Notes from systems thinking, history, design, and good fiction.

Evidence / 02

Side quests

Small experiments

Tools, sketches, and curious prototypes without a roadmap.

Evidence / 03

Elsewhere

Places + perspective

Street-level observations and the ideas that travel back to the work.

Zoom out. Follow the next board →

Case board / 04

Contact / Open Channel

Bring me the hard problem

For backend systems, production reliability, and agentic AI work, open a channel directly.

Evidence / 01

Email

Start a conversation

Best for project context, roles, and detailed engineering questions.

Open channel
Evidence / 02

LinkedIn

Professional profile

Experience, background, and a direct way to connect.

Open channel
Evidence / 03

GitHub

Code + experiments

Projects, prototypes, and the engineering trail behind the board.

Open channel

The board stays open.