AI · INFRASTRUCTURE · STREAMING
AI Workspace
Multi-Model LLM Application Platform with a provider-abstraction layer, enabling model switching without downstream code changes.
Ranjan Bhattarai / Senior Full-Stack Engineer
Engineering production systems around distributed architecture, event-driven infrastructure, payments, real-time applications, and AI-powered products.
System / Profile
ranjan.engineer
Core systems
Experience
5+
years engineering
Focus
AI
+ distributed systems
Event-driven services, Kafka pipelines, consumer groups, retry strategies and fault-tolerant workflows.
Production data systems, query optimization, indexing strategies and reliable persistence.
Low-latency communication systems with persistent state and volatile distributed state.
Production workloads and event-driven processing pipelines running on AWS.
LLM-powered applications with provider abstraction, structured outputs and streaming.
Telemetry, distributed tracing and evidence-driven incident investigation.
AI · Observability · Distributed Systems
An AI-grounded incident correlation and observability platform designed around telemetry, evidence and structured reasoning.
Architecture
Evidence-first incident pipeline
OpenTelemetry
Trace ingestion
Kafka
Event streaming backbone
Correlation
Telemetry relationships
Processing
Real-time analysis
Evidence Layer
Structured incident context
AI Reasoning
Evidence-first analysis
Incident Report
Automated post-mortem
AI · INFRASTRUCTURE · STREAMING
Multi-Model LLM Application Platform with a provider-abstraction layer, enabling model switching without downstream code changes.
E-COMMERCE · PAYMENTS · REAL-TIME
E-commerce platform backend with real-time inventory synchronization via Socket.io and secure webhook-driven Stripe payment and subscription handling.
CRM · REAL-TIME · BUSINESS SYSTEMS
Electronic Billing, CRM & Business Operations Platform. Owned cross-project implementation of CRM and real-time chat modules as independent Node.js/React services integrated with a core Laravel platform.
MACHINE LEARNING · COMPUTER VISION
A computer-vision side project using a PyTorch pipeline to detect cardamom leaf disease from real-world field photos.
Dec 2020 — Feb 2026
Unibird
Designed production backend systems spanning event-driven architecture, payments, real-time communication and AWS infrastructure.
Designed modular Node.js/TypeScript backend services and event-driven workflows using Apache Kafka with consumer groups, retry topics and dead-letter queues.
Built end-to-end Stripe payment workflows covering PaymentIntents, subscriptions and webhooks, moving synchronous checks toward asynchronous, idempotent processing.
Diagnosed PostgreSQL sequential scans with EXPLAIN ANALYZE, introduced targeted indexing and eliminated N+1 query patterns.
Built JWT-authenticated real-time chat using Socket.io, PostgreSQL as source of truth and Redis for volatile state.
Ran production services on AWS EC2 and built S3 + Lambda event-driven pipelines for CPU-intensive image processing.
I enjoy working on the difficult layer between an idea and a reliable production system. My work sits around backend architecture, distributed systems, real-time applications, payments, cloud infrastructure and AI.
node .............. online
postgres .......... online
kafka ............. online
aws ............... online
ai ................ exploring
06 / Contact
Have a system to build, a difficult engineering problem, or an interesting idea? Let's talk.
ranjanbhattarai659@gmail.com