Name: 95%+ Lower AI Costs: Building a Secure, Multi-Model Chat Platform on AWS
Start: 2025-09-23T18:00:00-07:00
End: 2025-09-23T20:00:00-07:00
Location: Amazon San Diego Tech Hub (Corp Office)

15 min presentation: **Message Router** presented by:
Tijuana AWS User Group
Said Ricardo Zúñiga Cortes
Ruben Stellios Ruiz Alonso
Llineth Maciel Mora
Hiram Zúñiga Lugo
https://github.com/utiliware/message-router

45 min presentation: **Building a Secure, Multi-Model Chat Platform on AWS**
Many companies raced to adopt GenAI through per seat tools and ended up with high costs, data risk, and duplicate pilots across teams. This talk tells the story of one enterprise that chose a different path. We replaced scattered subscriptions with an owned, governed platform on AWS that uses OpenWebUI for the chat experience and AWS Bedrock for managed models, with the option to add your own LLMs when it makes sense. The result was ninety five percent plus lower platform costs for this customer, faster delivery of real use cases, and clear visibility into usage and risk. Attendees will see the architecture, the steps to set it up, and the decisions that tied the technology to business outcomes. Savings vary by usage profile.

# **Join us to Discover:**

* How to design a reference architecture for a governed conversational AI platform using OpenWebUI with AWS services such as Bedrock, ECS or EKS, RDS, S3, and a vector store
* How to route across models and mix Bedrock with bring your own LLMs to balance reliability, latency, and cost
* Security and compliance patterns that satisfy enterprise requirements, including SSO to the corporate IdP, KMS encryption, private networking with controlled egress, content controls, and auditable logs
* Cost modeling and chargeback that replace scattered tool spend with centralized visibility and policy based budgets
* A short demo: multi-model chat in OpenWebUI with connectors to support ticket triage and account research workflows

# **Speaker Bio — [Aries Youssefian](https://www.linkedin.com/in/ariesyous/)**

Aries Youssefian is a Solutions Architect at VeUP with over fifteen years in enterprise IT. His background spans Dell and VMware along with OpenStack and private cloud startups, where he delivered production systems and guided customers through modernization. Today he focuses on consulting that brings governed GenAI platforms and data products to life on AWS. Aries helps enterprises replace ad hoc AI tools with a secure multi model chat platform using OpenWebUI and AWS services, with clear guardrails, cost control, and faster time to value.

This event is sponsored by **[VEUP](https://veup.com/)**

Mohit Gadkari

Dimple

Murty Chappidi

Amazon Web Services San Diego | Official Events

Technology

High Scalability Computing

Cloud Computing

Mobile Technology

Mobile Development

Knowledge Sharing

Big Data

Amazon Web Services

SaaS (Software as a Service)

DevOps

Web Services

IaaS (Infrastructure as a Service)

SQL Server

95%+ Lower AI Costs: Building a Secure, Multi-Model Chat Platform on AWS

Artificial Intelligence Applications

Amazon San Diego Tech Hub (Corp Office)

Share

Amazon Web Services San Diego | Official Events

95%+ Lower AI Costs: Building a Secure, Multi-Model Chat Platform on AWS

Amazon Web Services San Diego | Official Events

Details

Members are also interested in