Gaizka JiménezSoftware developerwith AI in production
Gaizka Jiménez is a software developer in Bilbao. He ships generative AI and mobile apps to production.
AI in Python, the product in TypeScript. The native app, all the way to the store.
Four years of software, two of them with RAG, agents and LLM pipelines in production. Until September 2026, AI and mobile engineer at CODX Digital. On the side I build emap, a Basque mobility map. It is on the App Store.
Jiménez
Short answers.
Who is Gaizka Jiménez?
Gaizka Jiménez is a software developer in Bilbao. He ships generative AI and mobile apps to production: the AI layer in Python, the product in TypeScript, and the native app as far as the store.
Is he available to join a team?
Yes. He is looking for a team in Bizkaia, on-site or hybrid, and also remote within Spain or the EU. Email is contacto@gaizkajimenez.com.
What is emap?
emap is a mobility map for the Basque Country: live public transport, door-to-door routing and open data. The iOS app is on the App Store and the web is at emapapp.com.
emap page→What is emap Labs?
emap Labs is an open benchmark for geographic retrieval in Basque and Spanish, with DOI 10.5281/zenodo.21282784. If the retriever invents an answer, it fails.
emap Labs page→
My own product, and team product.
They are in production: you can open them, download them or cite them. Each one made a different call (my own infra, several AI providers, native, or a closed ecosystem). That is the work.
- Work / 01SoloIn production
emap
Mobility platform for the Basque Country
RoleDesign, development and infrastructure — soloPublic transport across Bizkaia, Gipuzkoa and Álava with live waiting times, door-to-door routing by car, motorbike, bike, on foot and public transport, plus open-data layers. Published iOS app, public web and my own API. Trilingual ES/EU/EN with audited WCAG 2.1 AA accessibility.
Key callDepending on nobody else's routing API. I run three routing engines on my own VPS — OSRM for driving in the Basque Country, Valhalla for the rest of Spain and for motorbikes, OpenTripPlanner for transit — with blue-green deploys and rollback. When Madrid went live, the combined graph surfaced a frequencies NPE in OTP 2.7 whose upstream fix only exists in 2.8, which drops the REST API I rely on: I materialised the GTFS frequencies instead of migrating.
Results- iOS app published on the App Store (Navigation)
- CarPlay in production and native Android Auto via Car App Library
- 3D WebGL navigation with building occlusion and real heights from regional cadastres
- Live arrivals from 4 transport networks over GTFS-RT
React NativeExpoPythonFastAPIMapLibre GLOSRMOpenTripPlannerValhallaPeriod2026 — PresentRoleOwn project, soloOutputiOS shipped · web in production - Work / 02SoloPublished · Apache-2.0
emap Labs
Open benchmark for geographic retrieval in Basque and Spanish
RoleAuthor — Apache-2.0 code, CC-BY / ODbL dataThe data and AI layer behind emap, published with a DOI. A golden corpus of 139 hyperlocal searches across 22 layers, a strict held-out split and abstention cases: if the retriever invents an answer, it fails.
Key callThe finding I did not expect: the Basque gap was not about the language, it was about the model. With MiniLM the semantic stage barely contributed in Basque; switching to multilingual-e5-large raised held-out Spanish from 58% to 73% and Basque from 63% to 71%. The Basque in the corpus is written, then checked against Itzuli, the Basque Government's translator — not unreviewed machine translation.
Results- Citable DOI: 10.5281/zenodo.21282784
- 22 layers and 27,515 points of interest, versioned
- First hyperlocal mobility MCP server, with 5 tools
- Evals in CI with regression gates on every push
PythonFastAPIfastembede5-largeMCPGitHub ActionsPeriod2026RoleAuthorOutputOpen source · DOI · v0.2 - Work / 03TeamIn production
Kunoa
Health and nutrition with generative AI
RoleAI engineerAI backend for nutrition plans. A five-step generation pipeline on AWS Lambda, because long jobs blew past the API timeout. On top, several LLM providers and RAG over a nutrition knowledge base.
Key callLiteLLM with OpenAI and Anthropic Claude, so I am not stuck on one model: the provider is configuration, not architecture. Then RAG on Pinecone and photo-based food recognition to estimate macros.
Results- 35% churn reduction from AI personalisation
- Recognition base of 18,000+ dishes
- 5-step pipeline handling jobs beyond the timeout limit
PythonAWS LambdaLiteLLMPineconeSwiftKotlinPeriodCODX Digital · 2025–2026RoleAI and mobileOutputiOS + Android, in production - Work / 04TeamIn production
ChutApp
Social video for youth football
RoleMobile and backend engineerPlayers, teams and families share clips and training moments in a moderated environment built for under-18s. I worked on live streaming, the native clients and the moderation system.
Key callNative on iOS and Android from day one, no hybrid shortcut. Short video shows up in the gesture, the capture and the playback. You pay for that by shipping every feature twice.
Results- 12,000 users on the platform
- Live streaming in production with AWS IVS, operated at real events
- Content moderation system of my own design
SwiftKotlinNext.jsAWS IVSDynamoDBPeriodCODX Digital · 2025–2026RoleMobile and backendOutputiOS + Android, in production - Work / 05SoloBuilt · unpublished
China Survival Guide
WeChat Mini Program for foreigners in China
RoleSolo projectA guide for getting around China, shipped as a WeChat Mini Program. Two views, one content store, no backend. Built and tested inside WeChat's runtime.
Key callI stayed on WXML, WXSS and JavaScript. No bridge, no React, no backend. I wanted to see how WeChat works from the inside, not dress it up as something I already knew.
Results- Working build validated with 50+ expats
- Publication blocked only by the Chinese legal-entity requirement
WXMLWXSSJavaScriptWeixin DevtoolsPeriod2026RoleSoloOutputWorking · unpublished - Work / 06SoloIn review · App Store
Schengen Days
A 90/180 engine in your pocket
RoleDesign and development — soloOne rule: 90 days inside the previous 180. Travel history stays on the iPhone, no account, offline. React Native and Expo, same path as emap. Spanish and English. Apple has it in review: it is not downloadable yet.
Key callThe 90/180 count does not apply to me: I am an EU citizen. I saw the problem in Thailand. The app is the cut: an engine with tests against fixtures, not a passport index. Entry and exit count. If two trips overlap, the days merge.
Results- Rolling window from today back 179 days, with tests
- On-device data: no server to do the count
- Submitted to App Store Connect; no public listing yet
React NativeExpoTypeScriptPeriod2026RoleSoloOutputIn review · App Store
All of this can be verified.
A CV is read. This is opened: the app, the repo, the DOI, the evals. No form. No booking a slot with me.
- Proof / 01App Store
emap, published
Navigation app on the Spanish App Store. You can download it now, without asking me for anything.
View on the App Store↗ - Proof / 02Publication
DOI 10.5281/zenodo.21282784
Benchmark for geographic retrieval in Basque and Spanish, citable and Apache-2.0 licensed.
View the publication↗ - Proof / 03Code
github.com/r3tr0eth/emap-labs
Corpus, evaluation harness and versioned results. Reproducible in three commands, with evals in CI.
Open the repository↗ - Proof / 04Production
emapapp.com
Public web and API in production, on my own routing infrastructure on a VPS.
Open the site↗
English · updated 2026
I work so that whoever comes next understands it.
Habits I already use in the projects above. You can audit them in the repo.
Decisions get written down, not remembered.
I keep numbered architecture decisions (emap is up to D-023), with the context and the option I dropped. Whoever arrives later does not have to ask me why routing avoids a third-party API: it is written down, and dated.
A green CI is the definition of done.
A hard rule I hold myself to: after every push I check the pipeline, and if it is red that gets fixed before anything else. In the data layer the evals have regression gates: if a metric drops, the push does not pass.
No pretending.
What cannot be measured is left out. The benchmarks I publish reward an “I don't know” over inventing an answer. Limitations sit next to the results, coverage bias included. I would rather have one small number that survives a whole interview.
- Interface changes verified with a screenshot, not with a claim
- WCAG 2.1 AA accessibility audited with axe-core
- Spanish, Basque and English at parity, without machine translation
What's on the bench.
Smaller projects, in various states. I keep them visible: it is what I work on when nobody asked.
- Bench / 01Daily use
LanguageOS
My own language-learning system, built around immersion and context. Not gamified.
StatusInternal tool in daily use and active development.
language-os.vercel.app↗ - Bench / 02Open source
Susurro
Local voice dictation for Claude Code: transcription on your own machine, with no audio sent to any service.
StatusPublic code on GitHub, with a published build log.
github.com/r3tr0eth/susurro↗ - Bench / 03Waitlist
QuestClub
A private social game for groups of friends: seasonal missions and competition that lasts.
StatusOnboarding systems and gameplay loops in development.
questclub.app↗
Working notes on what I build.
The build logs cover how something got built, dead ends and decisions included. The rest are notes on AI, mobile and tech in Asia. Written in Spanish.
- 2026 · 09 · 07Field log
Play me rechazó por un aviso que solo estaba en la ficha
Google Play pidió tres cosas. Dos ya estaban en la ficha y en la web. La que tumbó el envío era la que no estaba dentro de la app: decir, en Ajustes, que emap no es oficial. De propina, un permiso de micrófono que no usábamos.
Read → - 2026 · 09 · 04Field log
Apple rechazó emap por un flag de audio que no usaba
Guideline 2.5.4. El revisor, en un iPad Air M4, no encontró audio persistente en segundo plano. Tenía razón: emap no es un reproductor. El flag lo había metido expo-audio por defecto.
Read → - 2026 · 09 · 02Field log
Play no me deja publicar hasta que 12 personas usen la app 14 días
El AAB de emap estaba listo. Google Play, con una cuenta personal nueva, no deja pedir producción hasta que 12 testers usen la app 14 días seguidos en closed testing. El reloj era el bloqueo, no el binario.
Read → - 2026 · 08 · 31Field log
He enviado mi segunda app a la App Store. Ahora toca esperar a Apple
emap ya se descarga. Schengen Days está en revisión. No me aplica la regla 90/180: el problema lo vi en Tailandia. Lo que he mandado a Apple es un motor, no un índice de pasaportes.
Read → - 2026 · 08 · 17Field log
Los peores bugs de emap no dieron ni un error
Repasando el registro de sesiones de emap encontré un patrón: los fallos que más tiempo me costaron no lanzaron ninguna excepción. Un service worker que dejó de repartir despliegues, una API caída devolviendo 200, props que la librería descarta sin avisar y el bug que tuve que arreglar dos veces.
Read →