Bilingual (Spanish/English) Full-Stack Developer
Minimum 5 years of professional software development experience. Full professional fluency in both Spanish and English — enough to lead technical discussions and produce documentation in both. This is essential, not a nice-to-have. Demonstrated full-stack experience with React, Python/FastAPI, and MongoDB or closely comparable technologies. Experience integrating with third-party platforms and systems : REST API design, data schemas, and data export/import workflows. Familiarity with conversational AI and LLM-based applications , including prompt engineering, retrieval-augmented generation, and vector search. Experience with cloud infrastructure (Google Cloud Platform preferred) and DevOps practices including infrastructure-as-code. Based in the Americas , or able to offer substantial working-hours overlap with Central America (GMT-6) . A track record of delivering high-quality software, ideally in resource-constrained or international development contexts. Ability to work independently and collaboratively in a small, distributed team, making pragmatic technical decisions under ambiguity while working directly with partners. Ability to explain technical concepts clearly to non-technical stakeholders, including chamber staff and government counterparts. The stack you'll work in: Frontend: React with Material-UI, TypeScript Backend: FastAPI (Python) Database: MongoDB with vector search AI/LLM: Google Gemini (and potentially open-weight models) for conversational AI Infrastructure: Google Cloud Platform (Cloud Run), Pulumi for infrastructure-as-code Integration targets: the chambers' labor intermediation platforms, via APIs and data export/import workflows Prior experience with these technologies or comparable stacks is required. You'll be the forward deployed engineer on this pilot: embedded with the partners, shipping in their environment, and an integrated member of Tabiya's dev team. Across roughly 40 days: Integrations (~12 days) — Build the CCIT and CCIC integrations with Tabiya's lead developer and the chamber-side IT firm: integration components against agreed API contracts and data schemas, data export/import workflows, connecting their platforms to our profiling, taxonomy and matching modules, and debugging in their environments. Configuration and Spanish localization (~10 days) — Own the standalone deployment for Choloma, and lead localization across all three chambers: UI, conversational content and prompts in Honduran Spanish, plus the occupational and skills taxonomy (ESCO mapping). Embedded technical liaison (~8 days) — Bridge the chambers' IT teams and platform providers (Spanish) with Tabiya's engineering and product teams (English): demos, technical working sessions, requirements, data schemas, and translating decisions and constraints both ways. Testing and QA (~5 days) — Integration and acceptance testing with chamber counselors and IT teams in Spanish, including review of conversational quality and skills extraction accuracy. Documentation and hand-off (~5 days) — Bilingual docs, deployment guides and training materials so the local IT firm can take over hosting and operation. You'll report to the World Bank task team on progress and deliverables; day-to-day coordination runs through Tabiya's engineering sprints. Success looks like: integrations live at CCIT and CCIC, a working standalone instance at CCICH, a tool that holds up in natural Honduran Spanish, and a hand-off complete enough that the chambers keep running it without you. Customer-facing engineering experience — forward deployed, solutions, or integration engineering — working directly inside partner or client environments and production systems. This is a strong advantage. Experience with multilingual application development and localization . Familiarity with Latin American labor markets or public employment services . Experience with open-source contribution workflows and public codebases. Prior work with international organizations, development agencies or the public sector . Familiarity with occupational and skills taxonomies (ESCO, O*NET or national equivalents).
Наблюдалась 2026-09-14, впервые 2026-09-11, источник — Get on Board (LATAM).