I've spent 5+ years writing production code for systems that can't break when it matters most — full stack development, event-driven backend architecture, and system design at enterprise scale. Still hands-on and shipping code every day.
One of my contributions improved processing performance by 40% — that came from rethinking the architecture, not just optimizing code.
I now design multi-agent AI systems — RAG pipelines, agentic orchestration, LLM integration — building them with the same rigor I apply to distributed systems: fault tolerance, idempotency, and correctness first, intelligence layered on top.
💻 Still hands-on daily — writing production code, not just designing on paper. Architecture decisions are backed by code I ship myself.
|
🏆 Recognition Top Performer — architecture contributions & measurable performance improvements |
🎓 Research Published work on ML-based heart disease prediction — where my applied AI journey started |
🟢 Open to: Senior Backend / Staff Engineer roles in FinTech · Payments · AI-native Platforms · Enterprise Automation · Distributed Systems
If you're building something hard, let's talk.
Production systems shipped across FinTech, gaming, blockchain, and enterprise workflow automation — each domain taught me something different about what "production-ready" actually means.
| 💰 FinTech | 🎮 Gaming | ⛓️ Blockchain | 🏢 Enterprise Automation |
|---|---|---|---|
| Financial automation & reconciliation at scale | Consumer product backends at scale | Hyperledger Fabric & Indy | Workflow & reconciliation systems |
Not just calling LLM APIs — designing the orchestration layer around them: how agents plan, retrieve, reason, and hand off work reliably.
🤖 Agentic AI & LLM Engineering
| Area | What I work with |
|---|---|
| 🧠 Agentic Orchestration | Multi-agent architectures, agent handoff & routing, task decomposition, LangGraph state machines |
| 🔍 RAG Systems | Retrieval pipelines, vector search, chunking strategies, grounding LLM output in real data |
| 🛠️ AI-Assisted Engineering | Claude Code, Cursor — used for architecture, refactors, and system design, not just autocomplete |
| ✍️ Prompt Engineering | Structured prompting, tool-use design, context management for multi-step agent reliability |
| 🔗 LLM Integration | Azure OpenAI, LangChain, LangGraph — wired into existing event-driven backend systems |
| 🏗️ System Design for AI | Agent systems built with distributed-systems rigor — fault tolerance, observability, predictable failure modes |
⚙️ Backend
🎨 Frontend
🗄️ Databases & Caching
☁️ Other — Infra, Cloud & DevOps
🔐 Security, Auth & Access Control
graph LR
A[Enterprise Data] -->|Event Stream| B[Kafka]
B --> C[Node.js + TS Microservices]
C --> D{Idempotent Processing}
D --> E[(PostgreSQL / MongoDB / Redis)]
C --> F[Multi-Agent Orchestration<br/>LangGraph]
F --> R[(RAG / Vector Retrieval)]
F --> H2[Azure OpenAI Agents]
H2 --> G[Automated Workflows]
D -.dead-letter queue.-> H[Fault Recovery]
style A fill:#26215C,color:#fff
style B fill:#000,color:#fff
style C fill:#0C447C,color:#fff
style D fill:#712B13,color:#fff
style E fill:#27500A,color:#fff
style F fill:#00D9FF,color:#000
style R fill:#534AB7,color:#fff
style H2 fill:#085041,color:#fff
style G fill:#085041,color:#fff
style H fill:#791F1F,color:#fff
Correctness, fault tolerance, and zero data loss first — intelligence layered on top, not bolted underneath.
🔥 I commit daily — every graph above pulls live from GitHub and updates automatically.
Open to Senior Backend / Staff Engineer roles at product companies working on hard problems in:
FinTech · Payments · AI-native Platforms · Enterprise Automation · Distributed Systems
