AI Project Builder

AI Project Builder

Generated

The biggest differentiator. Instead of 'build a portfolio,' OppZen recommends specific projects based on your target role, current skills, market demand, resume gaps, and interview weaknesses — each with a complete build spec.

Multi-Region Distributed Cache

Advanced4–6 weeks
Why this: Closes your top interview weakness (system design at scale) and your top skill gap (distributed systems).

A horizontally-scalable cache with replication, consistent hashing, and multi-region failover.

Show full project spec →
Learning Goals
  • Distributed Systems
  • Kubernetes
  • Consistency models
  • Observability
Tech Stack
GoRedisKubernetesTerraformPrometheus
Architecture
Sharded nodes with consistent hashing, gossip-based membership, async cross-region replication.
GitHub Structure
/cmd, /internal/cache, /internal/cluster, /deploy/k8s, /infra/terraform, /docs
Milestones
  • Single-node cache + tests
  • Consistent hashing ring
  • Replication
  • K8s deploy
  • Multi-region failover
Stretch Goals
  • TTL eviction policies
  • Client SDK
  • Chaos testing
Deployment
Managed Kubernetes across two regions via Terraform; GitHub Actions CI/CD.
Testing Plan
Unit + integration tests, load tests with k6, fault injection.
Documentation
Architecture decision records, README with diagrams, runbook.
Demo Suggestions
Live dashboard showing replication lag and failover; short Loom walkthrough.
Resume Bullet Examples
  • Built a multi-region distributed cache in Go handling 50k ops/s with sub-10ms p99 latency.
  • Designed consistent-hashing and gossip membership for zero-downtime node scaling.
Likely Interview Questions
  • How do you handle cache consistency across regions?
  • Walk me through your failover strategy.
  • How did you test for partition tolerance?
Recommended learning to build this
Linux Foundation logo
Kubernetes for Developers
for Kubernetes · Linux Foundation
O'Reilly logo
Designing Data-Intensive Applications
for Distributed Systems · O'Reilly
MIT OCW logo
MIT Distributed Systems (6.824)
for Distributed Systems · MIT OCW
CNCF logo
KubeCon Talks Playlist
for Kubernetes · CNCF
Stanford logo
Raft Consensus Paper
for Distributed Systems · Stanford
KillerCoda logo
KillerCoda K8s Labs
for Kubernetes · KillerCoda

Event-Driven Order Pipeline

Intermediate3–4 weeks
Why this: Builds your Kafka skill gap and demonstrates event-driven architecture for senior roles.

An order-processing pipeline using Kafka with idempotent consumers and a dead-letter queue.

Show full project spec →
Learning Goals
  • Kafka
  • Event-driven design
  • Idempotency
  • Observability
Tech Stack
JavaKafkaPostgreSQLDockerGrafana
Architecture
Producers → Kafka topics → idempotent consumers → outbox pattern → read store.
GitHub Structure
/services/order, /services/fulfillment, /libs/events, /deploy, /docs
Milestones
  • Topic design
  • Producer + consumer
  • Idempotency + DLQ
  • Outbox pattern
  • Dashboards
Stretch Goals
  • Exactly-once semantics
  • Schema registry
  • Saga orchestration
Deployment
Docker Compose locally; Kafka on managed cloud for the demo.
Testing Plan
Testcontainers for Kafka, contract tests, replay tests.
Documentation
Event catalog, sequence diagrams, README.
Demo Suggestions
Simulated order surge with live consumer lag and DLQ recovery.
Resume Bullet Examples
  • Designed an event-driven order pipeline with idempotent Kafka consumers and a dead-letter queue.
  • Implemented the outbox pattern to guarantee at-least-once delivery across services.
Likely Interview Questions
  • How do you guarantee idempotency?
  • When would you use exactly-once vs at-least-once?
  • How do you handle poison messages?
Recommended learning to build this
Apache logo
Kafka Documentation
for Kafka · Apache