I'm Rushabh

Architected a tree-shakeable TypeScript WebRTC SDK end-to-end as sole SDK engineer. Cut enterprise integration from 4–5 days to under 2 hours and delivered a 50-room race-condition-free state layer.

• ▸ Architected a tree-shakeable TypeScript WebRTC SDK with sub-200ms p50 latency; a plugin interface enabled 3 enterprise integrations without core SDK modifications.
• ▸ Eliminated state desync across 50 concurrent rooms by replacing last-write-wins Firestore with a server-timestamp reconciliation protocol, validated with a race-condition harness.
• ▸ Cut client integration from 4–5 days to under 2 hours as sole SDK engineer; aligned product and engineering on transport tradeoffs and delivered no-code onboarding end-to-end.

Launched a multi-tenant voice-agent platform serving 4 enterprises at 2.3K concurrent calls/day with 99.9% uptime, and redesigned the WebSocket turn-taking pipeline to cut TTFB to 800ms.

• ▸ Launched multi-tenant voice platform serving 4 enterprises at 2.3K concurrent calls/day with 99.9% uptime, automating 70% of manual SDR outreach workflows.
• ▸ Cut TTFB to 800ms by redesigning the WebSocket turn-taking pipeline with a custom VAD; deployed stateless GCP Cloud Run microservices with zero-downtime auto-scaling.

Co-architected a CAD-drawing validation platform under a semester-long engagement with ISGEC Heavy Engineering. Two-phase pipeline: synchronous rule engine + async LLM enhancement pass.

• ▸ Designed a two-phase pipeline: Phase 1 (parsing + 8-rule engine) completes in ~1.8s on a real 7.4MB drawing (~17.8K entities); Phase 2 runs as an async LLM enhancement pass with a hard timeout.
• ▸ Co-built an 8-rule geometric validation engine using Shapely spatial indexing (STRtree) for sub-quadratic clash detection, Hungarian-algorithm optimal assignment for 1:1 equipment matching, and RapidFuzz for fuzzy ID reconciliation — covered by 78 pytest cases at 91% statement coverage.

Industry-engagement project for Yantrikisoft: PyTorch STGNN with federated learning across retail silos, causal attribution via SHAP + DoWhy, and adaptive inventory replenishment via Ray RLlib. 14.2% → 11.1% MAPE across 4,200 SKUs.

• ▸ Improved 7-day MAPE from 14.2% → 11.1% across 4,200 SKUs via PyTorch STGNN; outperformed LightGBM by 2.1 and ARIMA by 3.1 MAPE points on held-out Q4 data.
• ▸ Fixed promotion misattribution via SHAP + DoWhy causal inference, directly informing enterprise campaign-budget allocation; cut cross-silo data transfer to zero using FedAvg federated learning.
• ▸ Replaced static inventory thresholds with adaptive reorder policies via Ray RLlib RL agents across multi-dimensional state-action spaces.