ARTIFICIAL INTELLIGENCE · ENERGY · INFRASTRUCTURE
AI SYSTEMS
FOR BUSINESSES AND
THE INFRASTRUCTURE
AHEAD
I build artificial intelligence systems, train teams and develop a vision that brings together computing, renewable energy and storage for the AI infrastructure ahead.
Daniel Brosed · AI systems architect
Huesca, Spain · AI, energy and infrastructure
ARTIFICIAL INTELLIGENCE · ENERGY · INFRASTRUCTURE
AI SYSTEMS
FOR BUSINESSES AND
THE INFRASTRUCTURE
AHEAD
I build artificial intelligence systems, train teams and develop a vision that brings together computing, renewable energy and storage for the AI infrastructure ahead.
Daniel Brosed · AI systems architect
Huesca, Spain · AI, energy and infrastructure
THE PROBLEM ISN'T YOUR TEAM
YOUR TEAM
REPEATS WORK
A SYSTEM SHOULD
BE DOING.
Chasing clients one by one, copying data between tools and answering the same questions every day isn't a lack of effort. It's a sign that a system is missing.
I build the part a machine can do better: your team keeps the judgement and the system takes on the repetitive work.
THE THREE PROBLEMS I SEE MOST
Manual prospecting
Finding, researching, qualifying and writing to every potential client without leaving a reusable process behind.
I SOLVE IT WITH
Automated prospecting and processes
Scattered processes
Data copied between tools that don't talk to each other, tasks that depend on one person and sales follow-ups that get lost.
I SOLVE IT WITH
Internal software built for your process
Unreviewed AI
Apps built in a hurry with exposed keys, misconfigured permissions or unprotected data.
I SOLVE IT WITH
Security audits for apps built with AI
I DON'T START WITH THE TOOL. I START WITH THE PROCESS THAT'S COSTING YOU THE MOST TIME, MONEY AND ENERGY.
THE PROBLEM ISN'T YOUR TEAM
YOUR TEAM
REPEATS WORK
A SYSTEM SHOULD
BE DOING.
Chasing clients one by one, copying data between tools and answering the same questions every day isn't a lack of effort. It's a sign that a system is missing.
I build the part a machine can do better: your team keeps the judgement and the system takes on the repetitive work.
THE THREE PROBLEMS I SEE MOST
Manual prospecting
Finding, researching, qualifying and writing to every potential client without leaving a reusable process behind.
I SOLVE IT WITH
Automated prospecting and processes
Scattered processes
Data copied between tools that don't talk to each other, tasks that depend on one person and sales follow-ups that get lost.
I SOLVE IT WITH
Internal software built for your process
Unreviewed AI
Apps built in a hurry with exposed keys, misconfigured permissions or unprotected data.
I SOLVE IT WITH
Security audits for apps built with AI
I DON'T START WITH THE TOOL. I START WITH THE PROCESS THAT'S COSTING YOU THE MOST TIME, MONEY AND ENERGY.
HOW I CAN HELP
THREE WAYS TO
WORK WITH AI.
I don't start by selling a tool or an off-the-shelf automation. I start by understanding which process is holding your team back, what data you have and what's worth building.
Then we agree on scope, timeline and price in writing. No subscriptions that almost fit, and no projects nobody knows how to maintain.
01
AI systems
I design and build agents, automations and internal software so that a repetitive process stops depending on one person, a spreadsheet or a chain of messages.
Process automation · AI agents · Internal software · Integrations
02
Team training
I train teams to use artificial intelligence with judgement in their daily work: with their own processes, their tools and cases they can apply the next day.
Hands-on sessions · Real cases · Prioritisation · Responsible use
03
Security for AI-built apps
I review apps built or accelerated with AI before they expose keys, data, permissions or access that shouldn't be open.
Authentication · Access · Data · OWASP
ONE SPECIFIC PIECE OR AN ENTIRE SYSTEM.
THE GOAL IS THE SAME: TECHNOLOGY THAT
SOLVES A REAL PROBLEM.
HOW I CAN HELP
THREE WAYS TO
WORK WITH AI.
I don't start by selling a tool or an off-the-shelf automation. I start by understanding which process is holding your team back, what data you have and what's worth building.
Then we agree on scope, timeline and price in writing. No subscriptions that almost fit, and no projects nobody knows how to maintain.
01
AI systems
I design and build agents, automations and internal software so that a repetitive process stops depending on one person, a spreadsheet or a chain of messages.
Process automation · AI agents · Internal software · Integrations
02
Team training
I train teams to use artificial intelligence with judgement in their daily work: with their own processes, their tools and cases they can apply the next day.
Hands-on sessions · Real cases · Prioritisation · Responsible use
03
Security for AI-built apps
I review apps built or accelerated with AI before they expose keys, data, permissions or access that shouldn't be open.
Authentication · Access · Data · OWASP
ONE SPECIFIC PIECE OR AN ENTIRE SYSTEM. THE GOAL IS THE SAME: TECHNOLOGY THAT SOLVES A REAL PROBLEM.
REAL CASES
CONCRETE RESULTS.
NOT PROMISES OF
“TRANSFORMATION”.
BEFORE
3 H 04 MIN
NOW
5 MIN
MARKETING AGENCY · AI SYSTEMS
Qualifying a new account:
from 3 h 04 min to 5 min.
Cesea Agenc.ia is a full-stack marketing agency. I design and build a layer of automation and AI agents to support prospecting, account research, opportunity qualification, sales follow-up and in-house production.
The team keeps its commercial and creative judgement. The system cuts the time spent researching, organising and preparing repetitive work.
29 FINDINGS
5 OF THEM SEVERE
DIGITAL PRODUCT · SECURITY AND AGENTS
Application security
and AI agents for Balority.
Balority is a social network for the startup ecosystem that connects founders, investors, mentors and talent.
I reviewed the application layer using the OWASP methodology and identified 29 security findings, five of them severe. I then built an internal AI agent workflow that helps find like-minded profiles within the network and makes connections without manual intervention.
Systems don't replace a team's judgement. They clear away the work that doesn't need a person to do it.
REAL CASES
CONCRETE RESULTS.
NOT PROMISES OF
“TRANSFORMATION”.
BEFORE
3 H 04 MIN
NOW
5 MIN
MARKETING AGENCY · AI SYSTEMS
Qualifying a new account: from 3 h 04 min to 5 min.
Cesea Agenc.ia is a full-stack marketing agency. I design and build a layer of automation and AI agents to support prospecting, account research, opportunity qualification, sales follow-up and in-house production.
The team keeps its commercial and creative judgement. The system cuts the time spent researching, organising and preparing repetitive work.
29 FINDINGS
5 OF THEM SEVERE
DIGITAL PRODUCT · SECURITY AND AGENTS
Application security and AI agents for Balority.
Balority is a social network for the startup ecosystem that connects founders, investors, mentors and talent.
I reviewed the application layer using the OWASP methodology and identified 29 security findings, five of them severe. I then built an internal AI agent workflow that helps find like-minded profiles within the network and makes connections without manual intervention.
Systems don't replace a team's judgement. They clear away the work that doesn't need a person to do it.
PROCESS
FROM THE FIRST CALL
TO A SYSTEM THAT WORKS.
Scope, phases, timeline and budget agreed in writing before we start.
01
We understand the process
You tell me which part of the business takes the most time, repeats the most or causes the most errors. Before talking about AI, we check whether automating it makes sense.
02
We define scope and proposal
You get in writing what we'll build, what's included, what's left out, the phases, the timeline and the budget. No “we'll see” and no surprise invoices.
03
I build and test
I design the system on top of your current tools whenever it makes sense. We test the workflow with real cases before putting it in the team's hands.
04
I deliver, document and train
You receive it working, documented and explained. If the team is going to use it, I train them so they can work with the system without depending on me for every change.
Tell me which process you want to improve
A first 30-minute conversation. No sales pitch, no commitment.
PROCESS
FROM THE FIRST CALL
TO A SYSTEM THAT WORKS.
Scope, phases, timeline and budget agreed in writing before we start.
01
We understand the process
You tell me which part of the business takes the most time, repeats the most or causes the most errors. Before talking about AI, we check whether automating it makes sense.
02
We define scope and proposal
You get in writing what we'll build, what's included, what's left out, the phases, the timeline and the budget. No “we'll see” and no surprise invoices.
03
I build and test
I design the system on top of your current tools whenever it makes sense. We test the workflow with real cases before putting it in the team's hands.
04
I deliver, document and train
You receive it working, documented and explained. If the team is going to use it, I train them so they can work with the system without depending on me for every change.
A first 30-minute conversation. No sales pitch, no commitment.
TRAINING
HANDS-ON TRAINING
FOR TEAMS THAT WANT TO
USE AI WITH JUDGEMENT.
AI is already within your team's reach. What's usually missing is knowing where it adds value, what the risks are and how to bring it into daily work without creating more noise than results.
I train teams with their own processes, tools and use cases. The goal isn't for them to depend on me: it's for them to know how to decide, build and use AI on their own.
FOR NON-TECHNICAL TEAMS
Using AI in daily work
Practical applications for writing, analysing, researching, organising information and designing useful workflows. We also learn to spot which tasks aren't worth automating.
FOR TECHNICAL TEAMS
Building systems on solid ground
Agent architecture, process automation, model integration, application security and criteria for taking a system from testing to production.
FOR LEADERSHIP
Deciding what deserves investment
How to spot real opportunities, prioritise projects, understand costs and risks, and decide where AI can generate a return without turning the company into a testing lab.
IT'S NOT A DEMO. IT'S WORK ON YOUR REALITY.
Sessions are prepared with context: which tools you use, which tasks repeat, what information you handle and what the team's goal is. Then we work on cases that can be applied the next day.
CREDENTIALS
Ongoing training in AI systems and ten certifications from Anthropic Academy, the official training of Anthropic, the company behind Claude.
Claude Code in Action · Introduction to Agent Skills · AI Fluency for Builders · Model Context Protocol: Advanced Topics
Do you want your team to stop trying AI at random and start using it with shared judgement?
Let's talk about training for your team
In-person or remote sessions, shaped around the team's level and processes.
TRAINING
HANDS-ON TRAINING FOR TEAMS THAT WANT TO USE AI WITH JUDGEMENT.
AI is already within your team's reach. What's usually missing is knowing where it adds value, what the risks are and how to bring it into daily work without creating more noise than results.
I train teams with their own processes, tools and use cases. The goal isn't for them to depend on me: it's for them to know how to decide, build and use AI on their own.
FOR NON-TECHNICAL TEAMS
Using AI in daily work
Practical applications for writing, analysing, researching, organising information and designing useful workflows. We also learn to spot which tasks aren't worth automating.
FOR TECHNICAL TEAMS
Building systems on solid ground
Agent architecture, process automation, model integration, application security and criteria for taking a system from testing to production.
FOR LEADERSHIP
Deciding what deserves investment
How to spot real opportunities, prioritise projects, understand costs and risks, and decide where AI can generate a return without turning the company into a testing lab.
IT'S NOT A DEMO. IT'S WORK ON YOUR REALITY.
Sessions are prepared with context: which tools you use, which tasks repeat, what information you handle and what the team's goal is. Then we work on cases that can be applied the next day.
CREDENTIALS
Ongoing training in AI systems and ten certifications from Anthropic Academy, the official training of Anthropic, the company behind Claude.
Claude Code in Action · Introduction to Agent Skills · AI Fluency for Builders · Model Context Protocol: Advanced Topics
Do you want your team to stop trying AI at random and start using it with shared judgement?
In-person or remote sessions, shaped around the team's level and processes.
VISION
AI NEEDS COMPUTING.
COMPUTING
NEEDS ENERGY.
The expansion of artificial intelligence doesn't depend only on models and software. It also depends on the energy available to train, deploy and run computing systems.
That's why my work brings together two lines that I develop separately today, but that are increasingly connected: AI systems and energy storage.
01
COMPUTING
AI systems need computing capacity, technical infrastructure and efficient use of the resources that sustain them.
02
ENERGY
Computing capacity needs stable energy. When the grid has limits, consumption and availability become part of the technological problem.
03
STORAGE
Renewables and batteries help to better manage when that energy is produced, stored and used.
IT'S NOT AN OPERATIONAL DATA CENTRE.
IT'S NOT A PROMISE.
IT'S THE STRATEGIC DIRECTION I'M WORKING TOWARDS.
ENERGY STORAGE FOR BUSINESSES.
VaultGrid Labs is my battery energy storage project for businesses. It initially focuses on Aragon and Catalonia and centres on understanding consumption, sizing a solution and coordinating the supply of the equipment.
It's a line independent from my AI consultancy, but part of the same direction: understanding how energy and computing will shape the systems we'll build in the coming years.
Specific information on energy storage on the VaultGrid Labs website.
VISION
AI NEEDS COMPUTING.
COMPUTING
NEEDS ENERGY.
The expansion of artificial intelligence doesn't depend only on models and software. It also depends on the energy available to train, deploy and run computing systems.
That's why my work brings together two lines that I develop separately today, but that are increasingly connected: AI systems and energy storage.
01
COMPUTING
AI systems need computing capacity, technical infrastructure and efficient use of the resources that sustain them.
02
ENERGY
Computing capacity needs stable energy. When the grid has limits, consumption and availability become part of the technological problem.
03
STORAGE
Renewables and batteries help to better manage when that energy is produced, stored and used.
IT'S NOT AN OPERATIONAL DATA CENTRE.
IT'S NOT A PROMISE.
IT'S THE STRATEGIC DIRECTION I'M WORKING TOWARDS.
ENERGY STORAGE FOR BUSINESSES.
VaultGrid Labs is my battery energy storage project for businesses. It initially focuses on Aragon and Catalonia and centres on understanding consumption, sizing a solution and coordinating the supply of the equipment.
It's a line independent from my AI consultancy, but part of the same direction: understanding how energy and computing will shape the systems we'll build in the coming years.
Specific information on energy storage on the VaultGrid Labs website.
Daniel Brosed
AI systems architect · Trainer ·
Founder of VaultGrid Labs
Huesca, Spain
ABOUT ME
I'VE BUILT SOFTWARE
SINCE I WAS 14.
BEFORE THIS, I WORKED
INSIDE A HOSPITAL.
Before dedicating myself to artificial intelligence systems, I worked as a nursing assistant, a diagnostic imaging technician and a radiotherapy and dosimetry technician.
I know what it's like to have to respond fast when there's little margin for error. I know what it's like to work with protocols, sensitive data and tools that have to work when they're needed.
Today I apply that same way of working to designing AI systems, automation, training and security: understanding the context before proposing a solution, documenting what matters and building with judgement.
I'm also developing VaultGrid Labs, an independent line of energy storage for businesses. It's part of a broader direction of work: understanding how computing, artificial intelligence and energy are going to need each other.
I'M NOT HERE TO FILL YOUR COMPANY WITH “AI”.
I'M HERE TO SEE IF THERE'S A REAL PROBLEM
A SYSTEM CAN SOLVE.
Daniel Brosed
AI systems architect · Trainer · Founder of VaultGrid Labs
Huesca, Spain
ABOUT ME
I'VE BUILT SOFTWARE
SINCE I WAS 14.
BEFORE THIS, I WORKED
INSIDE A HOSPITAL.
Before dedicating myself to artificial intelligence systems, I worked as a nursing assistant, a diagnostic imaging technician and a radiotherapy and dosimetry technician.
I know what it's like to have to respond fast when there's little margin for error. I know what it's like to work with protocols, sensitive data and tools that have to work when they're needed.
Today I apply that same way of working to designing AI systems, automation, training and security: understanding the context before proposing a solution, documenting what matters and building with judgement.
I'm also developing VaultGrid Labs, an independent line of energy storage for businesses. It's part of a broader direction of work: understanding how computing, artificial intelligence and energy are going to need each other.
I'M NOT HERE TO FILL YOUR COMPANY WITH “AI”. I'M HERE TO SEE IF THERE'S A REAL PROBLEM A SYSTEM CAN SOLVE.
CONTACT
IS THERE A PROCESS YOU
SHOULDN'T KEEP DOING BY HAND?
Tell me what's happening today, which tools you use and where the time is going. In a first conversation we'll see whether it makes sense to work together and what the next step could be.
If we move forward, you'll receive a proposal with scope, phases, timeline and budget agreed in writing.
PREFER TO TALK DIRECTLY?
You can book a 30-minute conversation or write to me on whichever channel suits you best.
Name and company
Email or phone
What would you like to talk about?
AI systems
Team training
Security audit
VaultGrid Labs / energy storage
Other
Tell me which process you want to improve
Controller: Daniel Brosed. Purpose: to handle your enquiry and reply to you. You can exercise your rights and find more information in the Privacy policy.
Daniel Brosed · AI systems architect, trainer and founder of VaultGrid Labs.
Huesca, Spain · Remote and on-site projects.
·
·
·
·
·
·
·
© 2026 Daniel Brosed · Handmade. No templates.
CONTACT
IS THERE A PROCESS YOU SHOULDN'T KEEP DOING BY HAND?
Tell me what's happening today, which tools you use and where the time is going. In a first conversation we'll see whether it makes sense to work together and what the next step could be.
If we move forward, you'll receive a proposal with scope, phases, timeline and budget agreed in writing.
PREFER TO TALK DIRECTLY?
You can book a 30-minute conversation or write to me on whichever channel suits you best.