LLM-Data Scientist

About the Company:

Cortex is an AI-powered platform built for real estate investors and asset managers. It transforms fragmented financial, technical, and commercial data into clear dashboards and predictive insights—helping real estate professionals optimize performance, reduce costs, and make smarter, faster decisions. Designed for strategic decision-making in real estate, Cortex uses advanced AI to make complex analysis and insights generation.

Responsibility

Multi-Agent & Orchestration Systems

Design and implement agent-based workflows using orchestration frameworks such as LangGraph. Develop reasoning pipelines, tool usage logic, and memory components as part of the platform’s AI architecture.

Structured Knowledge & Data Modeling

Contribute to the development of structured knowledge layers connecting financial and operational data to support AI-driven reasoning and analysis.

LLM Integration & Adaptation (API + Open Source

Integrate and evaluate both API-based and open-source LLMs (e.g., Llama, Mistral). Contribute to model selection, reliability improvements, and domain adaptation efforts. When relevant, participate in fine-tuning and optimization processes (e.g., LoRA / PEFT) to improve reasoning quality and alignment with product needs.

Context & Memory Integration

Implement mechanisms for context injection, retrieval, and memory usage (short-term and long-term) across agent workflows. Support grounding through structured queries, document retrieval, and knowledge lookups

Production Integration (Cloud Environment)

Work with engineering teams to integrate AI components into cloud-based infrastructure (GCP / Vertex AI), ensuring scalability, reliability, and performance.

Innovation & Applied Research

Stay up to date with advancements in agent-based systems, structured knowledge architectures, and open-source LLM development. Evaluate new approaches and contribute to integrating high-impact improvements into the platform.

Requirements

• 3+ years of experience building production-grade AI/ML systems in Python.
• Hands-on experience designing and deploying LLM-powered applications, including work with orchestration frameworks such as LangGraph (or similar), and integration of both API-based and open-source models.
• Experience working with open-source LLMs, including familiarity with adaptation or fine-tuning approaches such as LoRA / PEFT
• Strong understanding of retrieval, embeddings, context injection, and structured data integration across multiple data sources (e.g., structured databases, documents, and internal knowledge layers).
• Solid SQL and data modeling skills, with experience working with both structured and unstructured data.
• Experience deploying AI systems in cloud environments (GCP / Vertex AI preferred) and contributing to scalable, production-ready architectures.
• Strong system design mindset and ability to translate product requirements into robust AI components
• Demonstrated capacity for independent, hypothesis-driven problem-solving and translating ambiguous business questions into rigorous data science projects.



Nice to Have

• Experience building or contributing to multi-agent systems in production environments.
• Experience working with GCP-native services such as Firestore, Cloud Functions, or Vertex AI in AI-driven systems.
• Background in financial analytics, time-series modeling, or real estate data.
• Experience working with complex, multi-source data environments (structured databases, documents, internal memory systems).
• Understanding of data governance, privacy, and security best practices specific to sensitive financial data.