Machine Learning Engineer

About N71N71 is an AI intelligence platform built for companies that are tired of asking questions and waiting for answers. We connect a company's data — across CRMs, email, documents, and communication tools — into a unified knowledge graph, and use it to surface the insights that matter before anyone has to ask for them. We call it proactive intelligence, and it changes how teams operate.We're a small, high-conviction team building something that hasn't been built before. If you want to own a technical problem from first principles and see your work shape the product directly, this is the role.The RoleThe quality of what N71 surfaces to users lives or dies on retrieval. When our system decides what context is relevant to a user, what information to pull, and in what order to present it — that's your work.You will own the retrieval and ranking layer of N71's intelligence stack. That means building and improving the pipelines that take a user's context or query and return the most relevant, timely, and accurate information from across their company's data. You'll work closely with our knowledge graph infrastructure and the product layer to make the intelligence feel sharp and trustworthy.This is not a research role. We move fast, we ship, and we care about what works in production.What You'll Work OnDesign and build hybrid retrieval pipelines — combining dense vector search, sparse keyword retrieval, and graph traversal — for real-world business dataFine-tune and evaluate embedding models for domain-specific enterprise content (emails, CRM records, documents, Slack threads)Build ranking and re-ranking systems that account for recency, relevance, source reliability, and user contextCreate evaluation frameworks to measure retrieval quality systematically, so improvements are measurable and regressions are caught earlyOptimize latency and cost across retrieval at scale, without sacrificing qualityCollaborate with the product team to translate retrieval improvements into user-facing intelligence featuresWhat We're Looking For3+ years of hands-on experience building retrieval systems in production — RAG pipelines, semantic search, or recommendation systemsStrong fundamentals in information retrieval: you know why BM25 still matters and when dense retrieval falls shortExperience with vector databases (Pinecone, Weaviate, Qdrant, or similar) and embedding model evaluationComfort working with graph-structured data is a strong plusAn engineering mindset: you write clean, testable code and care about reliability as much as performanceLow ego, high ownership — you're comfortable working in ambiguity and making judgment calls with incomplete informationBonus PointsExperience building retrieval systems over multi-modal or heterogeneous enterprise dataFamiliarity with knowledge graph representations and how they complement vector retrievalPrior work at an early-stage startup or in a zero-to-one product environmentWhat We OfferDirect access to founders and real influence over product directionA focused, technically ambitious team with no bureaucracyCompetitive salary benchmarked to MENA market rates, with flexibility for the right candidateEquity at an early stageWe hire people who want to build something that matters 💜