Lead Software Engineer in Bengaluru. I design and operate distributed systems, real-time data platforms, and edge-AI infrastructure — mostly in Go, Rust, and gRPC.
Five years building robot-to-cloud platforms at Machani Robotics: low-latency telemetry, event-driven messaging over NATS and Zenoh, and sub-200ms model serving on constrained edge devices.
flowchart TB
root(("Skills.md"))
root --> be["Backend"]
root --> ds["Distributed Systems"]
root --> ai["Edge AI"]
root --> pf["Platform"]
be --> Go & Rust & gRPC["gRPC / MCP"] & PG["PostgreSQL"]
ds --> NATS["NATS.io"] & Zenoh & Redis["Redis Streams"] & Events["Event-driven"]
ai --> ONNX & Candle & GGUF & LoRA
pf --> Cloud["AWS / GCP"] & Terraform & Bazel & OTel["OpenTelemetry"]
classDef cat fill:#2A9D8F,stroke:#1E6B62,stroke-width:1.5px,color:#FFFFFF,font-weight:bold
classDef leaf fill:#F5EFE0,stroke:#D9B26B,stroke-width:1px,color:#3A3A3A
class be,ds,ai,pf cat
class Go,Rust,gRPC,PG,NATS,Zenoh,Redis,Events,ONNX,Candle,GGUF,LoRA,Cloud,Terraform,Bazel,OTel leaf
style root fill:#E0A458,stroke:#9C6B1E,stroke-width:2px,color:#1A1A1A,font-weight:bold
- gRPC-MCP Gateway — reflection-based bridge mapping gRPC services into MCP tool schemas, so LLM agents can call robotics backends directly. Type-safe, traced end-to-end with OpenTelemetry.
- Pulse — OpenTelemetry-native observability SDK for memory-constrained edge systems, with MCAP replay for synchronized multi-sensor analysis.



