Senior Applied AI Researcher
Software Engineering, Data Science · Full-time
Tel Aviv District, Israel · Tel Aviv-Yafo, Israel
About Jones
Jones is a fast-growing PropTech SaaS company transforming how insurance compliance is managed across the real estate and construction industries.
Our platform uses AI to automate complex insurance compliance workflows, combining document understanding, information extraction, validation, comparison, and business logic to deliver accurate and reliable decisions at scale.
We are looking for a talented Senior Applied AI / LLM Researcher to advance our AI capabilities, improve automation accuracy, reduce human intervention, and continuously optimize performance and operational costs.
The Role
This is an opportunity to take ownership of a significant AI challenge: transforming complex, human-driven business processes into highly accurate, reliable, and scalable AI-powered workflows.
You will research, design, and improve end-to-end AI agents capable of handling multiple stages of complex workflows, from understanding and extracting information from documents to interpreting requirements, validating information, comparing evidence, applying business logic, and making decisions.
Your mission will be to continuously improve the accuracy, reliability, and efficiency of these agents, enabling more processes to run autonomously with minimal human intervention.
A key part of the role is developing systematic evaluation and continuous improvement methodologies that identify AI failures, understand their root causes, and turn those findings into measurable improvements.
This is a hands-on, research-oriented role with strong end-to-end ownership and direct business impact. You will work closely with Engineering, Product, and domain experts to turn AI research into practical, production-ready solutions.
Key Responsibilities
- Lead the research and development of end-to-end AI agents that automate complex, multi-step business workflows.
- Design and improve AI pipelines covering document understanding, extraction, processing, validation, comparison, business logic, and decision-making.
- Research and implement advanced LLM-based approaches to improve reasoning, accuracy, consistency, and reliability across different stages of AI workflows.
- Design evaluation frameworks, benchmarks, and automated testing methodologies to measure performance at both individual-step and end-to-end workflow levels.
- Develop systematic approaches to identify, classify, and analyze AI failures, understand root causes, and continuously improve performance.
- Experiment with different LLMs, prompting strategies, model architectures, fine-tuning techniques, and orchestration approaches.
- Develop methods for handling ambiguous, missing, inconsistent, or conflicting information, including validation and uncertainty estimation.
- Reduce hallucinations, incorrect decisions, and unnecessary human intervention while maintaining high accuracy standards.
- Explore and implement approaches such as OCR, Vision Language Models, RAG, structured outputs, and post-processing techniques where they provide measurable value.
- Build and improve monitoring and feedback mechanisms to track AI performance in production, detect regressions, and support continuous optimization.
- Optimize AI workflows for accuracy, latency, scalability, and inference costs.
- Collaborate with Engineering, Product, and domain experts to integrate research outcomes into production systems and continuously improve automation rates.
Requirements
- 4+ years of experience in AI/ML engineering or applied research, with hands-on experience in LLMs, NLP, or Applied AI.
- Proven experience designing and implementing end-to-end AI-powered workflows or agents that automate complex business processes.
- Strong understanding of LLM capabilities, limitations, reasoning approaches, and practical techniques for improving accuracy and reliability.
- Experience building multi-step AI systems involving information extraction, processing, validation, reasoning, or decision-making.
- Proven experience designing evaluation frameworks, benchmarks, and systematic experimentation methodologies for AI systems.
- Experience analyzing AI failures, identifying root causes, and translating findings into measurable improvements.
- Practical experience with prompt engineering, LLM evaluation, and model optimization techniques.
- Experience developing production-grade AI solutions, including monitoring, performance optimization, and reliability improvements.
- Strong analytical and problem-solving skills, with the ability to translate complex research challenges into production-ready solutions.
- Ability to independently own complex technical challenges from research and experimentation through production deployment.
Nice to Have
- Experience with agentic AI systems, multi-step reasoning, or AI workflow orchestration.
- Experience with OCR, Document AI, or Intelligent Document Processing.
- Experience with Vision Language Models and multimodal AI.
- Experience with RAG, structured output generation, and data validation pipelines.
- Experience building automated evaluation, feedback, or continuous improvement systems for AI agents.
- Experience automating complex business processes in insurance, financial, legal, or compliance-related domains.
- Experience optimizing LLM inference costs and performance at scale.
- An MSc or PhD in Computer Science, Machine Learning, NLP, or a related field.
Why Join Jones?
- Own a meaningful AI challenge with direct impact on product performance and business efficiency.
- Work on complex, real-world AI automation problems that go far beyond document extraction.
- Build AI agents capable of performing sophisticated business workflows with increasing autonomy.
- Have significant influence over the company's AI capabilities and technical direction.
- Turn advanced AI research into production solutions used by customers every day.
- Join a collaborative, product-driven engineering organization.