Open Interview

Workbench shell

local-first
Provider not connected
Workbench/resume/projects/proj_44c7adddc8385d8e72a81720

Event Pipeline Rebuild

One project stays anchored as the deep-dive unit: context on the left, session history on the right, and explicit project prompts ready to practice.

project_id
proj_44c7adddc8385d8e72a81720
sessions
1
highlights
5
tech_stack
8

Project context

Project context stays visible because the deep-dive flow should practice storytelling, not generic chatting.

Rebuilt a synchronous order-processing pipeline into an event-driven architecture for higher throughput. / Built a personal interview prep workbench with grounded QA and structured review flows.

Highlights

Highlights give the deep-dive session concrete material to probe.

Led the migration from sync RPC to Kafka-driven events across checkout and fulfillment.
Added idempotency keys, DLQ replay, and tracing dashboards to reduce incident recovery time.
Open Interview
Designed parse-review-confirm workflow for interview notes and resume projects.
Implemented Next.js + TypeScript workbench routes with retrieval trace visibility.

Suggested deep-dive questions

Suggested prompts come from the structured project entity and seed the first session turns.

What problem was Event Pipeline Rebuild solving, and what part did you personally own?
Why did you choose Java and Spring Boot for Event Pipeline Rebuild?
Walk through the hardest part of Event Pipeline Rebuild and how you delivered "Led the migration from sync RPC to Kafka-driven events across checkout and fulfillment.".
What outcome, metric, or business impact proves Event Pipeline Rebuild worked well?

Tech stack

The stack is visible before the user starts answering follow-ups.

JavaSpring BootKafkaRedisClickHouseNext.jsTypeScriptSQLite

Session history

Session history stays beside the project, not hidden behind the generic QA route.