Home / Case studies

Agents we build and run in production.

Three real systems, described by what they do and how we keep them safe. We don't make up metrics or client quotes, and anything we could not verify is left out.

Client: Ricki Pollo (rickipollo.mx)

Restaurant ordering agent

How ByAgentX built a server-side order service for a live rotisserie restaurant: menu-validated orders written straight into the Poster POS, staff alerts, and humans accepting every order at the register.

Read the case study →
Operator: ByAgentX (internal infrastructure)

Private LLM infrastructure

How ByAgentX runs an open-weight DeepSeek model on four NVIDIA GB10 systems with tensor parallelism, as a private, fixed-cost inference layer for its own AI agents.

Read the case study →
Operator: ByAgentX and the businesses it runs

Multi-agent operations

How ByAgentX runs its own businesses with a team of AI agents: sales agents on web chat, Instagram and Facebook, an outbound voice and WhatsApp agent, and internal ops agents with shared memory, scheduled jobs and monitoring.

Read the case study →

Want this for your business?

Book a 20-minute call. We look at your channels, your systems and where an agent pays off, and you leave with a plan.