Businesses of every size, from a corner shop to an engineering firm. The industry changes; the way we work doesn't.
Artificial intelligence
Agents, automations, consulting, training and data, from the first use case to launch. The same team that builds it stays to measure and maintain it.
What we do
For a company that is just starting out or a group with years behind it: a service built around your challenges, with results you can measure from the first weeks.
Agents that serve and solve, process automations, assistants for your team and search over your documents. We build the whole product, from the AI engine to the screen where someone reviews and approves, and connect it to your systems, MCP servers included.
We find the use cases that pay off most, lay out a realistic roadmap and check compliance with the GDPR and the AI Act. First we tell you what to do; then, if you want, we build it.
We train leadership, business teams and developers in generative AI and agents, using cases from their day-to-day work. It is taught by the same engineers who build these systems.
Prediction models, computer vision and data engineering for what generative AI can't solve on its own. And the well-ordered data your agents need to answer well.
Method
We weigh each case by what it brings and what it costs to build, and agree with you on how success will be measured.
We choose the models, the technique (RAG, function calling, MCP) and the safeguards to suit your data and your constraints.
We test it with real cases and volumes to get a first measure of the return within weeks, before building it in full.
We connect it to your tools with a validation screen where your team has the final say.
We monitor answer quality every day and add safeguards against hallucinations, so it stays reliable.
We track the indicators set at the start and keep improving the agent: first we confirm the gains, then we extend them.
Why Bitgenio
We don't hand over a prototype and disappear: the same team puts it into production, watches over it and answers if something fails.
The same engineers who code agents and automations help you decide what to do first and train your team.
We work with Claude, ChatGPT, Mistral, DeepSeek or Qwen and with Amazon's and Google's platforms, and recommend what fits, not the latest thing.
Long relationships and clear conversations: we tell you what AI can do for your business and what it can't.
Work
Voice and generative AI · Life stories
A voice interview with AI that turns your memories into a book.
Writing your life story without writing a word. We built the web app and the API: the user answers by voice, the AI transcribes each answer and asks the next question, from the general to the personal, and at the end it writes the story in the first person and sends it laid out as a PDF. The script forbids inventing or embellishing anything the user hasn't said, and the interview can be paused and resumed with email reminders.
Document AI · Public tenders
Reading public tenders with AI: every requirement in the tender, with the page it comes from.
Icarus bids for public contracts whose tender documents are long and used to be read by hand, one by one. We built an AI tool that reads the tender documents and the rest of the file, extracts the full list of requirements with the page each one comes from, flags contradictions between documents and answers questions about the tender.
AI models deployed in secure environments
Hybrid teams · People and agents
A festival run with the support of AI agents.
Camarzius Fest is a music, history and food festival at the Roman Villa of Orpheus, in Camarzana de Tera (Zamora). Its operations live in an Obsidian knowledge base where people and agents work side by side: documentation, email support, the activity programme, contract review and the paperwork for public grants.
The festival app, for iOS and Android and in three languages, was built by a hybrid team: people and ten agents, each with its own role (functional and technical analysis, architecture, development, code review, QA, UX/UI and deployment).
AI agents · Market radar
An AI radar that scans a sector, scores what matters and turns it into content.
Twice a week, an agent reads the sector's news (Spanish and European press, business, product, marketing, technology and the month's reports), proposes the insights that pass the filter and scores them against five weighted criteria. Each one ends up in a note with the fact, its verified source and why it scored as it did. From Telegram you can add articles to the radar or ask for a draft. And the radar learns from what gets chosen and what gets discarded.
It works for scanning a market to find insights, for screening ideas before launching an initiative, your own or a client's, or for feeding your social media content.
Let's talk AI
We comply with European data protection and artificial intelligence regulations. All our AI projects are carried out in accordance with the General Data Protection Regulation (GDPR) and the EU Artificial Intelligence Act (AI Act).