Multimodal AI Engineer: Agentic Workflows & Production

Arabic.AI; United States; Remote; USD 120,000 - 180,000; Full time

Job summary

Tarjama is hiring an AI Engineer to design, develop, and deploy agentic and multimodal AI solutions leveraging language, vision, and speech. You will contribute to production-ready products by assisting in model integration, evaluation, optimization, and data pipeline prep.

You will collaborate with senior engineers, design robust workflows, implement tool-using agents, and ensure observability, logging, and cost-efficiency in production.

Qualifications

• Experience designing AI agentic workflows.
• Experience building multimodal AI systems.
• Experience delivering production-ready AI products.
• Knowledge of data pipelines and data quality.
• Ability to document and communicate AI development clearly.

Responsibilities

• Design and implement AI agentic workflows that orchestrate language, vision, and speech models for multi-step tasks.
• Build and extend task-driven, tool-using AI agents using modern agent frameworks and orchestration patterns.
• Implement decision logic, memory strategies (short-term and vector-based), and tool-calling mechanisms in production-ready systems.
• Improve agent reliability through structured prompting, planning strategies, and basic error-handling mechanisms.
• Develop and integrate text-based AI systems, including LLM pipelines, embeddings, rerankers, and RAG architectures.
• Build document understanding pipelines using OCR and vision-based models to extract structured insights from complex documents.
• Develop and optimize speech-based AI solutions, including STT, TTS, and conversational voice interfaces.
• Design and improve multilingual and translation pipelines with attention to performance and quality.
• AI Product Development & Deployment.
• Own the development of specific AI components or features from design through deployment.
• Deploy, monitor, and optimize AI systems in production with guidance on scalability and cost-efficiency.
• Collaborate with engineering teams to integrate AI components into maintainable and scalable systems.
• Implement logging, monitoring, and basic observability practices to track system performance.
• Model Evaluation & Optimization
• Define and apply evaluation metrics for AI systems, including accuracy, latency, and robustness.
• Benchmark and optimize models for inference performance and cost-efficiency.
• Run experiments and A/B tests to validate improvements and support data-driven decisions.
• Identify common failure modes and contribute to improving model robustness and reliability.

Skills

AI agentic workflows

Multimodal AI development

Data pipelines

Model evaluation & optimization

Production deployment

Experimentation & A/B testing

Cross-functional collaboration

Tools

LLM pipelines

RAG architectures

OCR

TTS

STT

Embeddings

Vector databases

Job description

Tarjama is hiring an AI Engineer to design, develop, and deploy agentic and multimodal AI solutions leveraging language, vision, and speech. You will contribute to production-ready products by assisting in model integration, evaluation, optimization, and data pipeline prep.

You will collaborate with senior engineers, design robust workflows, implement tool-using agents, and ensure observability, logging, and cost-efficiency in production.