Sambhav Sharma
Full Stack Engineer · Systems Architect
I design and scale high-performance systems, combining engineering, architecture, and research to drive measurable product impact.
I build distributed systems and high-scale applications with a focus on performance, reliability, and clean architecture.
At BOLD, I lead engineering efforts across system design, personalization platforms, and performance optimization - delivering systems used at scale.
I also explore system design through research and POCs, working at the intersection of engineering, product, and AI-driven systems.
I focus on building systems that scale without increasing complexity - prioritizing performance, reliability, and clear architecture over quick fixes.
Experience
- Built real-time personalization systems driving +22% user retention
- Reduced load times by 35% by migrating to optimized PWA architecture
- Designed event-driven systems using Kafka and Redis for analytics
- Developed reusable design systems → 25% faster development cycles
- Maintained 99.9% uptime through scalable cloud infrastructure
Built and scaled multi-platform applications across web, mobile, and desktop
- Built cross-platform apps using React, React Native, and Electron
- Improved performance by 45%, significantly increasing user engagement
- Designed scalable CMS with role-based access and real-time updates
- Led frontend architecture and UX improvements across products
- Managed and mentored a team of 8–12 engineers
Systems & Architecture
- Designed microservices for high-scale applications using Kafka, Redis, and containerized infrastructure
- Built real-time data pipelines powering analytics and personalization
- Led POCs for AI-driven experimentation and recommendation systems
- Exploring LLM-based systems for product intelligence and automation
- Defined reusable architectural patterns across teams
Research
- Studied how resume tools perform across ATS systems and design platforms
- Analyzed global user behavior and tool preferences
- Built evaluation pipelines to measure ATS parsing performance
- Identified trade-offs between system trust, design control, and outcomes
- Proposed AI-driven systems (LLMs, explainable scoring, dual-mode editing)
Case Study
AI Personalization System · BOLD
Improving user engagement through real-time personalization
Problem
Low user engagement and retention due to generic content delivery across high-traffic platforms.
Approach
Designed a real-time personalization system using event-driven architecture to dynamically adapt user experiences based on behavior.
Architecture
- Event ingestion via Kafka for user activity streams
- Real-time processing and caching using Redis
- Personalization logic layer for dynamic content decisions
- A/B experimentation framework for continuous optimization
Trade-offs
- Chose eventual consistency over strict consistency for scalability
- Balanced latency vs accuracy in real-time recommendations
- Optimized for system throughput over heavy computation models
Impact
- +22% increase in user retention
- +30% increase in feature engagement
- Improved responsiveness of personalization systems