cienciadedades.cat logo

AI engineering built for impact

We are a consulting AI engineering business specializing in agentic systems, hyper-performant search, and LLM optimization to turn complex workflows into measurable outcomes.

Services

Consulting and delivery for teams that want agentic systems, search, and AI products that ship fast.

Agentic systems engineering
Design and build production-grade agents with the OpenAI Responses API and the Anthropic Agent SDK, plus tool orchestration, memory, and evals.
Learn more
Hyper-performant search
Hybrid retrieval, reranking, and caching for fast, grounded answers with on-prem options.
Learn more
LLM fine-tuning & evaluation
Customize models for your domain and validate them with rigorous benchmarks and regression tests.
Learn more
AI product integration
Ship AI features across web and iOS with secure APIs, observability, and cost-aware scaling.
Learn more
Claude Code training
Training programs that boost programming productivity with Claude Code and proven workflows.
Learn more
Custom AI consulting
Need something specific? We tailor the roadmap, architecture, and delivery to your business.
Learn more

Technology stack

Models, agent tooling, infrastructure, and privacy-first deployment for production-grade AI.

Models and agent tooling

OpenAI Responses API, Anthropic Agent SDK, and open-source stacks for model training and evaluation.

OpenAI
OpenAI
Anthropic
Anthropic
Hugging Face
Hugging Face
PyTorch
PyTorch

Inference and infrastructure

Hyperfast inference providers like Groq and Cerebras, plus cloud infrastructure for scale.

Groq
Groq
Cerebras
Cerebras
Azure
Azure
Vercel
Vercel

Partners and deployment

Microsoft partner delivery, iOS development, and privacy-preserving on-prem LLM deployments.

Microsoft Partner
Microsoft Partner
iOS Development
iOS Development
On-prem LLMs
On-prem LLMs

Selected projects

Examples of how we help organizations apply AI to real operations.

  • 01

    Cartel detection system

    Automated detection of anti-competitive behavior by analyzing public procurement pricing data.

  • 02

    Hospital stay forecasting

    Predictive model to optimize hospital resources and improve patient care delivery.

  • 03

    RAG assistant for internal documentation

    Built a virtual assistant that uses retrieval to answer questions on internal knowledge.

  • 04

    Specialized content generation

    Tuned an LLM to generate high-precision legal content with domain terminology accuracy.

Blog

Notes from the field on agentic systems, search, and AI delivery.

Local LLM agentic app for the Barcelona Supercomputing Center
Project update • More details coming soon

We built a local LLM agentic app for the Barcelona Supercomputing Center. We'll share the technical architecture and learnings in upcoming posts.

Let's talk about your project

We help you move from idea to production with AI systems that are reliable and measurable.

Email us

Reach out with questions, ideas, or partnership opportunities.

hola@solucionsdedades.cat

Book a call

Schedule a session to scope your project and next steps.

Calendly
Barcelona, Spain