Backend & AI Engineer building scalable distributed systems, event-driven pipelines, and production-grade LLM infrastructure.

I specialize in designing and operating distributed backend systems that handle real-world production workloads reliably and efficiently.

Currently working as an AI Engineer at Insurance Samadhan, where I:

• Architect Kafka-based event-driven microservices
• Build production LLM + RAG pipelines
• Optimize backend systems for throughput and latency
• Design fault-tolerant async processing workflows
• Implement observability and centralized logging infrastructure

Key Impact:

• Increased system throughput 3×
• Reduced backend latency by 56%
• Lowered infrastructure costs by 92%
• Improved document extraction accuracy by 40%

Distributed Systems

• Kafka, Redis, Event-Driven Architecture
• Idempotent Consumers, Dead-Letter Queues
• Async Processing & Concurrency Control
• Horizontal Scaling & Caching Strategies

AI Systems Engineering

• Production LLM Applications
• RAG Pipelines & Embeddings
• Vector Databases
• AI Workflow Optimization

Backend Engineering

• Python, Node.js, TypeScript
• FastAPI, Django, Express.js
• REST, gRPC, WebSockets
• MongoDB, PostgreSQL

Observability & Infrastructure

• Prometheus, Grafana
• Centralized Logging Systems
• Docker, AWS, GCP

A cryptographically verifiable, time-sortable distributed identifier system designed for high-concurrency environments.

• 18-byte compact format
• HMAC-SHA256 signature validation
• Monotonic sequencing with deterministic conflict resolution
• Benchmarked with 10M+ parallel ID generations (zero collisions)

Repository: https://github.com/KishanKumar08/veridjs

• Design for failure from day one
• Optimize for performance and cost
• Build systems that scale predictably
• Keep architecture clean and observable
• Ship AI systems that work in production, not demos

LinkedIn: https://www.linkedin.com/in/kishan-kumar08/
Email: kmali4551@gmail.com