Open to new opportunities · View experience · Get in touch
New York Metropolitan Area · Open to new roles
Distributed systems
rigor, applied to AI.
I bring distributed systems rigor to AI — the correctness guarantees, the operational discipline, the infrastructure thinking. Not an ML researcher. Not a platform admin. I build the rails under intelligence.
Multi-agent RAG system shipped. Building a distributed task queue with Raft-lite quorum replication, and an Agent+MCP Platform — MCP became a Linux Foundation standard in Dec 2025, and I wanted to understand it at implementation depth.
5+
years shipping
10M+
txns / day
300M+
records processed
30%
infra cost cut
<2%
hallucination rate
What I do
full experience →Distributed Systems
Kafka EOS · CQRS · Event Sourcing · Raft consensus · exactly-once at 10M txns/day
Cloud Architecture
100TB+ migration · AWS EMR · multi-AZ · Terraform · 30% cost reduction
Data Pipelines
30+ Airflow DAGs · Spark on EMR · 300M+ records/day · 99.5% SLA
CI/CD & Platform
Jenkins → GitHub Actions · OIDC · 20+ repos · 3× release velocity
Observability
OpenTelemetry · Splunk · Dynatrace · Prometheus · 50% MTTR reduction
AI Infrastructure
RAG in prod · LangGraph · pgvector · <2% hallucination · MCP
Writing Recent posts
all posts →Projects Featured work
all projects →Multi-Agent RAG Orchestration
A 4-agent retrieval-augmented system on LangGraph with RAGAS evaluation. Faithfulness 73% → 91% over a single-chain baseline.
Distributed Task Queue with Replicated State Store
Fault-tolerant task scheduler with 3-node quorum replication and exactly-once execution under node crashes. Workers auto-scale on queue depth via Kubernetes HPA.
Distributed Social Media Platform
A federated social network built from first principles — multi-node Go backend, Raft consensus for leader election, CRDTs for feed merging, and a timeline fan-out system that stays consistent under network partitions.
"The cleanest production system is the one whose failures are obvious."
From a kafka post-mortem, 2024