Mario Rios / GenAI Systems
Production-focused GenAI systems on AWS and Azure
I build LLM products that ship: multi-agent workflows, RAG pipelines, evaluation, and reusable platform services. Focus on measurable outcomes, cost control, and reliability in production.
Relocating to Munich in July 2026; Chancenkarte valid from July 15. No sponsorship required.
Highlights
Latency
~84s / 1s
Audio processed per second of compute.
Scale
100K+
Documents automated per month with RAG.
Recognition
Top 5%
Credicorp AI Programme (120 participants).
Case studies
Sea Sentinel AI
Upstream and downstream data pipelines with model benchmarking for classification, AWS automations on DynamoDB and AI services, and end-to-end onboarding of external data providers.
Munich Hybrid · Germany
AWS DynamoDB and AI service automations
E2E Data provider onboarding, cost to deploy
Credicorp
Multi-agent credit-risk decision support on Azure AI Foundry plus a production-ready multi-tenant GenAI content platform for 4 business units.
85%
4 agents
4 BU
Entel Perú
RAG automation for 100K+ documents per month, 84s-to-1s audio inference, LLM evaluation, and AWS serverless AI infrastructure.
100K+
~84s / 1s
+25%
Main stack
Core
Supporting
Core
Supporting
Core
Supporting
Professional profile
AI Platform / Software Engineer with five years building production systems across finance, retail, and telecom. Recent work centers on GenAI platforms, evaluation pipelines, and multi-agent orchestration for business-critical use cases.
Experience
5 years
Backend, data, and cloud infrastructure.
Industries
3 industries
Finance, retail, and telecoms.
Education
UTEC / BSc CS
CodeBERT-SMCH thesis (secondary).
Credentials
Languages
Certifications
Microsoft Certified: Azure AI Fundamentals
Microsoft
Issued Feb 2026
Credential ID F1E5FE4C352C151
Cambridge B2 First
Score: 168
Cybersecurity Essentials
Cisco