AI Engineer

DREAMVISION - first senior AI engineering lead.We are building creative intelligence: private, versioned systems that encode how exceptional creators and creative organisations perceive, reference, judge, refuse and direct creative work.Our foundational pipeline is already in place. We are now looking for a highly capable AI engineer to turn it into a secure, testable and deployable product.The RoleYou will work directly with the founder to build the technical architecture behind the first Creative Genomes and early studio deployments.Your focus will span the full path from composition to runtime:Structuring tacit creative judgementRepresenting references, criteria, refusals and decision logicEvaluating fidelity to the creator or organisationManaging confidence, uncertainty and human escalationVersioning the Genome as the practice evolvesDeploying the validated intelligence through the DREAMVISION Genome APIIntegrating the runtime into existing internal studio and company workflowsThis is an early-stage role with significant responsibility and influence over the technical foundations of the company.What you will buildSystems that transform structured conversations, project readings, comparison exercises, references, corrections and decision evidence into usable reasoning material. The final architecture may combine structured schemas, retrieval, graph representations, language models and deterministic systems.The method should fit the problem. We are not interested in adding technical complexity for effect.Fidelity SystemsCreative judgement cannot be evaluated through conventional benchmark accuracy alone.You will help develop methods for assessing:Fidelity to validated creative judgementConsistency across comparable casesQuality of reasoningAppropriate refusalConfidence calibrationCorrect human escalationDrift between Genome versionsDisagreement between the system and its creatorPrivate APIYou will help turn the runtime into a private, versioned API that can integrate with existing internal systems, review environments and proprietary studio tools. The API must return structured creative intelligence, not simply unstructured model responses.What we are looking forYou have likely built AI systems where the difficult part was not calling a model, but designing the architecture, context, evaluation and controls around it.You should be highly capable in:PythonProduction backend engineeringLLM application architectureStructured extraction and schema designRetrieval and knowledge systemsReliable structured outputsHuman-in-the-loop systemsEvaluation of subjective or weakly labelled domainsConfidence, refusal and escalation logicAPI designData provenance and versioningObservability and failure analysisPrivacy-conscious system designExperience in some of the following would be particularly valuable:Knowledge graphs or graph-based reasoningMulti-stage or agentic systemsEmbeddings and semantic retrievalFine-tuning or preference optimisationMultimodal AIModel routingOntology and taxonomy designTemporal or versioned data modelsEnterprise integrationsSecure isolated deploymentsCreative software, design technology or computational designWhat we valueWe are not looking for someone who treats a larger prompt as an architecture.We value:Explicit representations over hidden assumptionsEvaluation over demosTraceable reasoning over plausible languageHuman escalation over reckless autonomyCareful systems over technical theatreSimple solutions when simple solutions are sufficientThe role requires research judgement, but its purpose is to build and ship a real product.How we workDREAMVISION is an early-stage company working at the intersection of creative intelligence, reasoning systems and the future infrastructure of creative practice. The team is intentionally small.Engineering, research and product development are closely connected. You will work directly with the founder, early creators and selected studio partners rather than through layers of management.We care about precision, intellectual honesty and careful execution.This role is notA prompt-engineering positionA conventional chatbot projectA generative-content roleA pure academic research postA narrowly scoped backend positionA role maintaining an already settled architectureYou will be helping define the architecture itself.Location: Dubai / remote with meaningful overlap across UAE and European working hoursType: Full-timeStage: Founding teamReporting to: Founder and CEOHow to applyPlease send:Your CV or LinkedIn profileGitHub, portfolio or selected technical workA short note explaining why this problem interests youAn example of an AI system you have designed or taken into productionA brief description of the hardest architectural decision you made while building it