EJ.Let’s talk

Software engineer · Toronto, Canada

Distributed systems. Developer infrastructure. Measurable results.

I’m Eesher Janda, a Computer Science student who builds and validates reliable software—from Raft consensus and CRDTs to CI intelligence and real-time products.

Explore selected work GitHub
3.8 / 4.0GPA · University of Guelph
375KCI observations processed
1,000seeded Raft fault scenarios
85%internship test coverage

01 / Selected work

Proof over promises.

Each project starts with a difficult failure mode and ends with reproducible evidence—not just a feature list.

01Developer infrastructure

TriageCI

A self-hosted CI failure-intelligence service that turns noisy test failures into explainable flaky-test and regression signals.

  • HMAC-authenticated, idempotent ingestion
  • Bounded worker queue with backpressure
  • Transactional SQLite WAL storage
  • Prometheus metrics and GitHub webhooks

Measured result

12,261observations / second

375,000 observations processed across three 64-client stress runs with zero processing failures.

TypeScriptNode.jsSQLiteDocker
Read the engineering breakdown
02Distributed systems

RaftKV

A three-node, crash-recoverable key-value store built from scratch to make Raft’s hardest failure modes observable and testable.

  • Majority commits and linearizable reads
  • Conflict-hint log repair and snapshots
  • C11/JNI write-ahead log with CRC32
  • Torn-tail recovery and fault injection

Measured result

1,000seeded fault scenarios

21,621 committed writes, 36,992 blocked RPCs, 17,000+ duplicate deliveries, and a verified 186 ms failover.

JavaC11JNIPythonDocker
Read the engineering breakdown
03Local-first software

SyncLab

An encrypted collaborative editor that keeps accepting edits offline, converges after reconnect, and keeps document contents opaque to the relay.

  • Operation-based CRDT and anti-entropy
  • Persistent encrypted outbox replay
  • Ordered ASP.NET Core WebSocket relay
  • Out-of-order and duplicate delivery tests

Measured result

250Kreplica-operation checks

Validated convergence over 200 randomized trials with five replicas, 20% temporary loss, and 10% duplicates.

C#ASP.NET CoreReactTypeScript
Read the engineering breakdown

02 / Experience

Shipped in the real world.

Codentrel

Software Engineering Intern

Faster decisions, fewer manual reports.

Built and shipped a React analytics dashboard for 50+ KPIs, replacing recurring Excel aggregation for internal analysts.

2×report throughput
8 → 18 req/s
50%lower P95 latency
12s → 6s
<500mslive refresh
via WebSockets
8defects caught
before production

03 / About

Curious about the edge cases.

I study Computer Science at the University of Guelph, where I’ve earned a 3.8 GPA and Dean’s Honour List recognition. I’m most drawn to software where correctness, performance, and clear reasoning all matter.

Outside class, I help with Google Developer Groups workshops and hackathons, then put the same build–test–explain loop into my own systems projects.

01

Distributed systems

Raft consensus, CRDTs, replication, recovery, linearizability, and failure-oriented design.

02

Backend & data

Node.js, ASP.NET Core, REST, WebSockets, PostgreSQL, SQLite, Redis, indexing, and observability.

03

Delivery & quality

Docker, Linux, GitHub Actions, fault injection, load testing, Jest, Vitest, and clear engineering docs.

04 / Contact

Have a hard problem worth solving?

I’d love to hear about the system, the constraints, and what success looks like.

Email me LinkedIn GitHub