Capellaris
Your company with dedicated and trainable AI employees. Multi-tenant agent platform on FastAPI, React and GCP, with a pilot enterprise customer live in production.
Barcelona, Spain — open to new engagements
$
Daniel Scheidemantel Camargo. Eight years designing, shipping and operating scalable, high-traffic systems on AWS and GCP — production LLM-agnostic agent platforms, cloud-native modernisation, and the data pipelines underneath them.
AI agent management platform
A platform for managing AI agents specialised in different business areas. Agents communicate with one another, form squads dynamically, and keep learning — turning scattered automation into an organisational capability.
Marketing, finance and legal agents built on a shared orchestration layer.
CrewAI and MCP tool-calling over a unified multi-LLM abstraction.
Agents assembled on demand around the task that needs solving.
Hybrid retrieval over pgvector and BM25, with prompt versioning and evaluation.
Your company with dedicated and trainable AI employees. Multi-tenant agent platform on FastAPI, React and GCP, with a pilot enterprise customer live in production.
Open source energy consumption and environmental monitoring for data center infrastructure, built on low-cost hardware and published across nine papers.
Consolidation and High Availability on Virtualized Environments — reducing energy consumption in cloud infrastructure while keeping services available.
Open source projects, experiments and contributions.
GreenHop: Open source PUE continuous monitoring for small and medium data centers
MeHarCEn: A Method of Harmonizing Energy Consumption in Data Centers
CHAVE: Consolidation with High Availability on Virtualized Environments
Free · no commitment
Bring an architecture you want reviewed, an AI use case you are unsure how to productionise, or a legacy platform that needs modernising. You leave with a concrete opinion either way.
Replies within one working day