United States · From Barcelona
AI development company for the US.
Custom AI development for startups and companies in the United States: LLM applications (built on language models such as GPT or Claude), AI agents, RAG, machine learning models and AI MVPs, from a senior engineering team that has been shipping software since 2009. The whole team is in Barcelona and works in English with clients across the US, with hours that overlap US time zones.
Talk to the team- 15+ yearsshipping software from Barcelona
- 100%of calls scored in real time for a US contact center
- Since 2017the same engineering team behind TAO Group Hospitality’s platform
One cell.
End to end.
The same senior team from start to finish. No hand-offs between departments.
AlamedaDev is an AI-native product engineering studio founded in 2009 in Barcelona. We build web, desktop and mobile products with AI at the core, specialised machine learning (speech-to-text, document AI, scoring, ranking, forecasting) and new ML and LLM features for products already in production. An AI cell is a small senior team of software engineering, applied ML and product that owns the full cycle, from proof of concept to running it in production.
- DiscoverWhich problem, with what data, and how success is measured.
- DesignArchitecture, interface and where a person decides.
- BuildShort iterations in your repository and your cloud.
- EvaluateQuality, cost per use and latency on real cases.
- DeliverCode, documentation and handover to your team.
Selected work.
Contact center
AI in an existing product · US (under NDA)
Speech to text tuned to the client’s vocabulary and supervised scoring models on AWS Kinesis that score 100% of calls in real time, for compliance and sales potential.
TAO Group Hospitality
Web and mobile product · Since 2017
Operations, ticketing and loyalty on one platform across every venue, built and maintained by the same engineering team since 2017.
Meter
Web product · Networks
Real-time IoT dashboards for company networks across many sites, built for support staff, field technicians and customers.
Reedy
Desktop app · AI · macOS
An AI meeting copilot for the Mac (usereedy.com): a briefing before the call, live suggestions during it and a searchable transcript after.
What we build.
- AI integrationLLM and ML features inside the product you already run in production, with evaluation and cost control.See service
- Machine learningCustom models for scoring, ranking, forecasting and speech to text, trained on your data.See service
- AI MVPsThe first version of an AI product for startups, ready for real users and built to keep growing.See service
- AI agentsAgents that do multi-step work across your tools, with evaluation and human approval.See service
- RAG and Document AIAnswers over your documents that cite their source, with per-user permissions.See service
- Web and desktop appsThe full product around the AI, or plain software when AI isn’t needed.See service
Work in production.
TAO Group Hospitality, Meter, Reedy and more than twenty companies run software we designed and built.
See the workIn Barcelona.
The whole team is in Barcelona. From here we work in English with startups and companies across the US, with meetings in your time slot and written updates in between.
ContactThe basics.
Do you have an office in the US?
No. AlamedaDev was founded in 2009 in Barcelona and the whole team works from there, in English, with startups and companies across the US.
How do you handle the time difference?
We set overlap windows with your time zone (Eastern or Pacific) for reviews and decisions; the rest moves forward through written updates in English. Every project has a clear technical lead.
Do you build AI MVPs for startups?
Yes. We scope the smallest version that tests the idea with real users, build it with code meant for production, and hand over the repository and documentation so your team can keep going.
Who leads the AI work?
Ryan Porter, Chief Scientist and Lead AI Solutions Architect, with 20+ years in enterprise software. He has led R&D labs in the US that built and scaled ML models, LLMs and multi-agent systems.
Where is the data hosted?
In your AWS, Google Cloud or Azure account and the region you choose, for example a US region. If the data cannot leave your infrastructure, we use open models deployed inside it.
What is an AI cell?
A dedicated senior team of software engineering, applied ML and product that owns the full cycle: discovery, architecture, development, evaluation and handover to your team, with the code and documentation in your repository.