Travel Review Platform: Moderation, Fraud, and the Recency Problem
Running a travel review platform at scale — review eligibility, fraud detection, ranking that resists gaming, and the moderation workflow that makes it trustworthy.
Senior Software Engineer · Dhaka, Bangladesh
Designing resilient
distributed
systems.
Hi, I’m Rahiyan Safin. I build scalable, event-driven backend platforms with Java, Spring Boot, and C#/.NET. I help teams turn complex enterprise constraints into dependable software.
Fig. 01 A system is only as good as its recovery path.
Java & Spring Boot.NET & C#Distributed systemsEvent-driven architectureCloud engineeringDatabase internals
Featured articleOne problem.
One problem.
All the trade-offs.
Long-form engineering notes for readers who care about what happens between the architecture diagram and a running enterprise system.
CockroachDB from First Principles: The Database That Refuses to Die
Why a global SQL database is really a problem of agreement and time, how Google solved it with TrueTime, GPS, and atomic clocks, and how CockroachDB reaches a practical design on ordinary infrastructure.
inside
Read the complete article →- CockroachDB
- Architecture
- Distributed Systems
Latest articles
Browse everything →Notes from the workbench.
How Async and Await Work in C#
A story-led explanation of Tasks, compiler-generated state machines, continuations, threads, I/O completion, exceptions, cancellation, contexts, and practical async design in .NET.
How the ASP.NET Core Runtime Handles a Request
A plain-English walkthrough of how an ASP.NET Core application starts, builds its middleware pipeline, accepts traffic through Kestrel, selects an endpoint, and returns a response.
Portfolio intelligenceSkip the scrolling.
Skip the scrolling.
Ask a precise question.
Search my articles, background, skills, and contribution history in plain English. Every result takes you back to something you can read and verify.
EXAMPLE QUESTION How did you move consumers without losing offsets?
RS✦
FROM THE ARTICLETreat the target offsets as a new coordinate system.
Start each Event Hubs consumer group at an approved position, reconcile with stable event IDs and business sequence numbers, and keep Kafka available until the rollback window closes.
Read the article →Selected impactProblems that
Problems that
needed deciding.
Private sessions run for product teams. The companies are under NDA, so each entry gives the problem and the decision rather than the client. Full write-ups are in workshops.
Event-driven reliability
A team could publish but could not reason about what happened after a consumer crashed mid-batch. Rebuilt the failure story around idempotency keys, a poison-message path, and a replay operators had actually rehearsed.
KafkaIdempotencyReplay
Platform migration
Four brokers, no capacity plan, and a broker upgrade that ate a weekend. Worked through what changes when the platform runs the cluster: compatibility gaps, partition mapping, dual running, cutover, and the rollback nobody had tested.
KafkaAzure Event HubsMigration
Change data capture
Two systems that had to agree, maintained by hand, forever. Replaced the nightly job with logical replication into a stream, then spent most of the session on the hard part: schema evolution and what a consumer does when it receives a shape it has never seen.
CDCDebeziumSchema evolution
Runtime performance
Built for a team whose services passed load tests and still timed out in production. Read a state machine out loud, traced a continuation onto a thread pool, and found where a request was blocking something that looked asynchronous.
C#AsyncThreading
Payments security
A third-party callback was the only part of the flow nobody owned. Drew the trust boundaries, moved the secret handling, made the audit trail answerable, and scoped what genuinely had to live inside the cardholder data environment.
SecurityPaymentsAudit
Selected contributions
Explore the full organization network →Organizations I have built for.
Binary Quest
- Strategic Tech Innovators
TAP
38organizations across product delivery and private engagements
SSW Consulting
SVT Robotics
Shine Solutions
Neolution
JTL Software
NovoTX
Eximia
SOFTURE
Client work is described at a safe level. No private repositories, confidential metrics, or NDA material are published here.
A little context
I design for scale, failure, and change.
Dependable enterprise systems feel predictable because their boundaries, failure modes, observability, and recovery paths were designed deliberately.
I am a senior software engineer based in Dhaka with 8+ years of experience across enterprise SaaS, manufacturing, fintech, and healthcare. My core work spans Java/Spring Boot, C#/.NET, event-driven architecture, and cloud platforms.
More about me