Company DescriptionNafes.ai is the first tender analytics and bid management platform in the Arab world. Combining advanced artificial intelligence and data science, Nafes.ai enhances the success and predictability of bids. The platform empowers organizations to unlock revenue from tenders and drive scalable growth, offering innovative solutions tailored to the competitive bidding process.The RoleWe're looking for an AI Engineer to own and advance the brain of our product. You'll be the person who makes Nafes smart — responsible for LLM orchestration, retrieval systems, document intelligence, and the AI features that set us apart from every other tender platform in the market.This is a hands-on builder role. You'll write production code daily, ship features end-to-end, and own the AI services our entire product depends on.What You'll OwnDesign and maintain our LLM service layer, prompt engineering, model selection and routing, fallback logic, response caching, and cost optimization across leading AI providersOwn the embedding and search infrastructure powering tender matching, BOQ similarity, and competitor discoveryBuild and improve systems that extract structured intelligence from Arabic government procurement documents, classification, entity recognition, and compliance detectionImplement the AI logic behind our core product features: Go/No-Go analysis, win probability scoring, estimated tender cost, risk assessment, competitor intelligence, and smart recommendationsMonitor and optimize AI API spend, embedding efficiency, and reliability at scaleWhat We're Looking For3+ years building AI/ML systems in production, not just research or notebooksStrong Python with hands-on async framework experience (FastAPI or equivalent)Proven experience integrating and orchestrating LLM APIs in shipped productsWorking knowledge of vector search and embeddingsExperience with document processing, PDF extraction, OCR, or structured data extraction from unstructured sourcesAbility to ship clean, documented, production-grade code with proper error handlingComfortable owning a domain end-to-end with minimal supervision in a startup environmentStrong PreferencesArabic language proficiency, our entire data domain is Arabic government procurementExperience with Arabic NLP, tokenization, entity recognition, text classificationBackground in procurement, tenders, or government contractingTrack record of AI cost optimization, prompt caching, model routing, embedding deduplicationExperience building recommendation or scoring systems in productionWhat Success Looks LikeIn your first 30 days, you'll be onboarded, shipping your first production improvement, and documenting the AI systems you own. By 90 days, you'll have full ownership of our LLM orchestration layer with at least one measurable improvement to cost efficiency or model accuracy. Within 6 months, you'll be driving AI roadmap items, recommendation agents, confidence engines, pricing intelligence, from design through deployment. You'll be the go-to person for anything AI at Nafes.