Markdown
Backend Systems Engineer | Go Β· Rust Β· Python Β· Distributed Systems & Media Infrastructure
Building high-concurrency backend infrastructure, low-latency video streaming pipelines, and P2P networking systems.
- π Current Focus: Building open-source Rust toolkits for RTSP ingest and P2P camera streaming (targeting India's CCTV market shifts).
- ποΈ Background: Joined as one of the first two engineers at Ease My AI, building the backend infrastructure of RDX (a production CCTV/VMS platform handling 10,000+ cameras) from zero and leading a team of 10.
- β‘ Core Specialties: Distributed systems, high-concurrency streaming pipelines, real-time protocols (WebRTC, RTSP, QUIC), and memory-safe systems programming.
| Category | Technologies / Skills |
|---|---|
| Languages | Go Rust Python |
| Media & Streaming | go2rtc, MediaMTX, ZLMediaKit, RTSP, WebRTC, ONVIF, fMP4 |
| Networking & P2P | QUIC, NAT Traversal, STUN/TURN/ICE, TCP/UDP, WebSocket |
| Backend & Architecture | Microservices, RESTful APIs, System Design, Distributed Systems, Event-Driven Architecture |
| Databases & DevOps | Docker / Docker Swarm, Kubernetes, Redis, MongoDB, Linux, Git, Kong Gateway |
| Leadership | Technical Planning, Code Review, Team Management (10+ devs) |
π¦ camproto β Open-Source P2P Camera Streaming SDK
Rust-based camera infrastructure tools targeting modern replacement demands for CCTV markets.
- Tech Stack: Rust, Tokio, WebRTC, RTSP, QUIC, ONVIF, fMP4
- Highlights:
- RTSP ingest layer parsing H.264/H.265 streams with
custompackage. - GOP-aligned fMP4 segment storage with binary seek indexes.
- P2P transport implementation using WebRTC with STUN/TURN NAT traversal.
- RTSP ingest layer parsing H.264/H.265 streams with
Engineered core Video Management System handling high-concurrency CCTV streams for 10,000+ cameras.
- Tech Stack: Go, Python, Docker Swarm, Redis, MongoDB, ZLMediaKit, go2rtc, Kong Gateway
- Highlights:
- Scaled backend infrastructure from zero to production as one of the first two core engineers, leading a team of 10.
- Integrated
go2rtc,MediaMTX, andZLMediaKitfor high-concurrency stream handling in a distributed Docker Swarm environment. - Built a distributed camera assignment coordinator utilizing Redis leader lease for High Availability (HA).
Live web application and signaling architecture demonstrating zero-intermediary media routing.
- Tech Stack: Rust (
axum), WebRTC, coturn, Upstash Redis, Go, Svelte, Cloudflare Pages - Highlights:
- Rust signaling server featuring role-based WebSocket relay and HMAC-SHA1 time-limited TURN credentials.
- Network topology detection logic (local network vs. mobile data routing).
- Portfolio: nitinkalma.dev
- LinkedIn: linkedin.com/in/nitin-kalmaste-04a69489
- Email:
kalmastenitin@gmail.com