Software Engineer • Distributed Systems • Backend • AI Systems • MSCS @ University of Georgia, Athens, GA (4.0 GPA)

Links: 🌐 LinkedIn • 🧑‍💻 LeetCode

I build high-performance distributed systems, AI-driven automation, and scalable backend platforms.

My work spans:

• ⚡ Backend Engineering (Python/FastAPI, Java/Spring Boot, gRPC)
• 🕸️ Distributed Systems (distributed key-value store, Kafka, consensus, replication, anti-entropy)
• 🗃️ Databases (PostgreSQL, Redis, Neo4j, MongoDB)
• 🧠 AI Systems & LLM Engineering (RAG, agents, embeddings)
• ☁️ Cloud Infra + Observability (Docker, Prometheus, Grafana, AWS/GCP)

I am skilled at translating complex technical challenges into clean, reliable, and production-ready systems.

Aug 2025 – Present Built a distributed AI grading platform that processes 1,000+ concurrent jobs with:

• 🧩 RAG + Agentic workflows → Eliminated manual effort entirely (100% automation)
• ⚡ 70% latency reduction (8 → 2 minutes) with async persistence & Redis caching
• 🔐 SSO + OAuth2, HMAC idempotency, rate limiting, and secure pipelines
• 🛰️ gRPC multi-provider LLM gateway → 90% fewer schema errors
• 📊 Full observability: Prometheus, Grafana, p95/p99, dashboards

Mar 2024 – Jul 2024

• ⚙️ Designed scheduling/ETA engines (FastAPI + PostgreSQL) → 98% latency reduction
• 🤖 Built a RAG Invoice-to-JSON Agent (GPT + Tesseract + Pinecone) → 100% automation, <1.5s latency
• 🚛 Production vehicle-routing engine → boosted deliveries 20%, scaled to 5× traffic
• 🧵 Introduced Redis caching + LRU → cut external API usage 20%

Jan 2023 – Feb 2024

• Built a 5,000+ DAU real-time messaging platform (Node.js + Angular + Kafka + MongoDB)
• Implemented Kafka microservices → 95th percentile latency −30%, throughput 3×
• Rebuilt mobile app (Flutter + Hive) → 3× faster load time; 50% onboarding drop reduction
• Engineered a GPT-3 counselor bot → cut support load/costs 50%

Spring Boot, Neo4j, Kafka, Elasticsearch, AWS

• 3k read RPS / 1.2k write RPS | p95 <220 ms | p99 <350 ms (10k VUs)
• Added tracing, metrics, request IDs, canary deploys, automated rollbacks (<2 min)

Java, gRPC, Distributed Systems

• Adaptive quorum selection → GET p99 −30% under stragglers
• Repair-aware anti-entropy → 50% fewer bytes, 50% faster convergence
• Chaos-tested with Toxiproxy & netem

FastAPI, React, Redis, Claude 4.5

• Streaming backend → time-to-first-claim 1.2s → 300ms
• Retrieval + LLM reranking → precision 0.70 → 0.85
• Privacy-by-design (no DB, structured logs, replay-safe streams)

• Languages: Python, Java, C/C++, TypeScript, JavaScript
• Backend / Systems: FastAPI, Spring Boot, gRPC, Node.js, Express
• Databases: PostgreSQL, MongoDB, Redis, ElasticSearch, Kafka, Neo4j
• DevOps: Docker, Kubernetes, GitHub Actions (CI/CD), Prometheus, Grafana
• Cloud: AWS, GCP, Azure
• AI / LLM: LangChain, Llama, Pinecone
• Other: Linux, Bash, Airflow

• 🧩 700+ LeetCode problems solved (Top 5%)
• 🧑‍🏫 TA for Software Engineering - mentored 80+ students as a Graduate TA (Software Design Patterns, Git, CI/CD).
• 🏆 Awarded full scholarship + stipend @ UGA for excellent academic performance
• 📚 Research in time-series forecasting (Transformers, VAR, CNN)
• 🎇 Open-Source contribution in:
  • ScalaTion & ScalaTionPy - A data science library written in Scala and Python
  • Dynamic-Sarimax - Developer & Maintainer for a time-series package
• 📈 Developed a Proprietary Trading Algorithm yielding up to 50% returns in a quarter.
• 🔭 Currently exploring LLMs and systems engineering