⚡ Building production AI systems that don't just demo — they scale.
class SumitKumarGupta:
title = "Senior Software Engineer | Applied AI Lead"
location = "Hyderabad, India 🇮🇳"
experience = "6+ years in production systems"
current_focus = [
"Multi-Agent LLM Orchestration",
"RAG Pipelines & Document Intelligence",
"Tax Compliance Automation at Scale",
"LLM Cost Optimization & Evaluation",
]
origin_story = """
Started as a React / React Native engineer.
Got obsessed with AI. Rebuilt myself from the ground up.
Now shipping multi-agent systems that process 2M+ fields/month.
"""
fun_fact = "I cut LLM costs 35% ($15K/month) while improving accuracy. That's my sport."| Metric | Achievement |
|---|---|
| 🤖 Queries Automated | 50K+ support & filing queries/month via AI agents |
| 🧾 Fields Processed | 2M+ monthly via OCR + LLM pipeline |
| 🎯 Extraction Accuracy | 95%+ on unstructured financial documents |
| 💸 Cost Reduced | 35% LLM cost savings → $15K/month |
| ⚡ Latency Improved | p95 from 850ms → 700ms |
| 📋 Tax Filings Automated | 10K+ with 100% audit compliance |
| 🩺 Healthcare Consultations | 10K+ with real-time AI summarization |
| 📉 Manual Entry Reduced | 90% reduction via intelligent automation |
🐍 Backend & Infrastructure
⚛️ Frontend & Mobile (former primary expertise)
🔧 Dev Tooling & Practices
🏗️ Currently @ Zenwork (AI Lead)
└─ Multi-agent tax copilot: 50K+ queries/month
└─ LLM + OCR pipeline: 10K+ forms, 95% accuracy
└─ Transitioning 100% manual tax workflows → 80% autonomous
📚 Learning
└─ Advanced agentic evaluation frameworks
└─ LLM fine-tuning & RLHF
└─ Distributed inference optimization
🤝 Open to collaborate on
└─ Production LLM systems with real-world impact
└─ AI x Fintech or AI x Healthcare
└─ Open-source RAG / agent tooling
2019 ──── Software Engineer @ Unthinkable Solutions
↳ Anomaly detection | 500GB/month telemetry | 0.92 F1-score
↳ 1M+ point data viz with Web Workers (95% latency reduction)
2022 ──── Senior Frontend Engineer @ Aptedge
↳ Global search widget | 50K queries/month | <200ms p95
2023 ──── Senior Full Stack Engineer @ ARTHINK AI
↳ AI-powered healthcare platform | 10K+ consultations
↳ HIPAA + SOC 2 compliance | Azure OpenAI integration
2025 ──── SDE-II / AI Lead @ Zenwork ◀ NOW
↳ Multi-agent LLM orchestration at scale
↳ $15K/month cost savings | 95% extraction accuracy
↳ Org-wide AI evaluation framework
I'm always happy to discuss:
- 🔬 RAG architecture — chunking strategies, hybrid retrieval, re-ranking
- 🤖 Multi-agent systems — orchestration, tool-calling, failure recovery
- 💰 LLM cost optimization — caching, routing, prompt compression
- 📄 Document intelligence — OCR pipelines, structured extraction, validation
- 🏥 AI in regulated domains — HIPAA, tax compliance, audit trails