Piyush Bhuyan
Applied AI @ Keyloop | Building Agentic Systems & AI Infrastructure
Hyderabad, India
works.piyushb@gmail.com
I build AI systems . I work as an Applied AI Engineer (aka forward-deployed engineer aka FDE). At work, it can range from anything like creating a multi-agent system (Claude do not make any mistakes) or spending hours studying legacy code and questioning the architecture. Outside work, I usually experiment with new models, read papers, or catch up with the local community.
Applied AI Engineer
Keyloop
- Architecting autonomous incident orchestration across 38 enterprise products; 80% of events routed at $0 LLM cost through deterministic rules.
- Reduced P2 MTTR by 99% (90 min → under 90 sec) and major-incident resolution time by 97% (5h 53m → under 10 min).
- Building a generative AI pipeline that converts incident patterns into validated runbooks in under 7 days, with automated GitHub PRs.
Associate Engineer - AI Enablement
Keyloop
- Built an AI operations platform handling 200 req/s and 500+ concurrent engineers; improved throughput 40×.
- Cut critical-path database latency 620× and shift-handover generation time 93%, saving 30+ engineering hours/month.
- Built an ML-powered infrastructure lookup service saving 1,000+ engineering hours annually.
Software Development Intern
Keyloop
- Built real-time infrastructure dashboards across 10,000+ servers, improving performance review cycles by 25% and saving 6+ engineering hours weekly.
- Automated Salesforce-based SLA reporting with Python ETL and connection pooling, saving 48+ engineering hours annually.
Built an open-source Python static-analysis CLI that detects hallucinated imports in AI-generated code before runtime. Six analysis rules, JSON/SARIF reporting, and CI/pre-commit integration.
Built a multi-tenant AI support platform using FastAPI, LangGraph, and RAG, with persistent agent memory, RBAC, tenant isolation, and LLM observability.
Layered Security - Intrusion Detection
Published a layered Naive Bayes + XGBoost intrusion-detection system with GAN-based augmentation; achieved an 80.20% F1-score on NSL-KDD. Springer CIS 2025.
Bachelor of Engineering
Vasavi College of Engineering · 9.04
Machine Learning Operations Engineer Associate (AI-300)
Microsoft
Sept 2026Certified SAFe 6 Practitioner
Scaled Agile, Inc
Mar 2026Certified Cloud Practitioner (CLF-C02)
AWS
Feb 2026
Attendee · Oct 24, 2026