Final-year CS at RGIPT. I build backend and applied-AI systems in Go, Python and C++; I use Zed btw.
$ go test -bench=. github.com/Adityarya11/...
strata-kv durable writes, batched fsync 346 → 45,867 rec/s
strata-kv missing-key lookups, bloom filters ~100× faster
strata-kv startup, 25 segments 825 ms → 2.9 ms
var end of speech → first audio (p50) 2.07 s on a 4 GB laptop GPU
var LLM time to first token (p50) 0.28 s
hiero merged PRs (python 30 · c++ 15 · go 3) 48
PASS- StrataKV: embedded key-value store in Go. Checksummed WAL, atomic batches, Bloom filters. On pkg.go.dev.
- Voice Agent Runtime: a voice agent that runs on one laptop. WebRTC → Go orchestrator → Whisper, Qwen and Piper over gRPC.
- Blan: a small language. C++17 interpreter, Go job queue, and a playground you can try.
- Hiero SDKs: committer on the Python SDK, Linux Foundation Decentralized Trust.
Now: teaching my voice agent to stop talking when you interrupt it, and looking for 2027 backend / applied-AI roles.
adityarya.vercel.app · LinkedIn · X · arya050411@gmail.com