Software Engineering student building production-grade backend systems, learning AI/ML, and shipping real products from scratch.
- π€ Building AI agents, pipelines, and intelligent applications
- ποΈ Implementing FAANG-level system designs with full test suites
- π Thinking in architecture before thinking in code
- π Open source contributor β Oppia, EduAid, Resonate
Implementing FAANG system designs as fully-tested Python codebases β proving I understand the infrastructure AI runs on.
| Project | Core Concept | Deep-Dive Metric |
|---|---|---|
| ποΈ blob-storage | S3/MinIO internals | XOR erasure coding, 3Γ replication, SHA-256 chunk integrity |
| β±οΈ rate-limiter-cluster | Distributed rate limiting | Atomic Lua scripts β no token-sync race conditions across Redis nodes |
| π·οΈ web-crawler | Scalable crawler | Priority frontier, politeness delays, cryptographic content dedup |
| π search-autocomplete | Typeahead at scale | In-memory Trie, map-reduce trending, sub-30ms via CDN edge cache |
| π¬ video-transcoder | HLS media pipeline | 10MB chunks, concurrent async workers, broker-tracked state |
| π¬ chat-backbone | WhatsApp/Discord scale | WebSocket hub, heartbeat pruning, offline message queues |
| π notification-router | Async multi-channel | Exponential backoff + jitter, dead-letter queues |
| π¦ distributed-cache | In-memory KV store | O(1) LRU/LFU, stampede shield, consistent-hash sharding |
| π analytics-ingestion | Mixpanel pipeline | Kafka partitioning, Redis spike absorption, batch flushing |
| π distributed-id-generator | Twitter Snowflake | 4,194,304 IDs/ms across 1,024 workers, clock-backward safe |
| Project | Description | Contribution |
|---|---|---|
| oppia/oppia | Free online learning platform (Google-backed) | Active contributor 2026βpresent |
| AOSSIE-Org/EduAid | AI-powered quiz generation | PyTorch compatibility fixes, dependency resolution |
| AOSSIE-Org/Resonate-Website | Open source video platform | Fixed React memory leak in cleanup lifecycle |

