Duxre · 1 week ago
Senior/Lead AI Developer
Duxre is a rapidly growing, AI-first startup poised to redefine the CRE industry. The role is a senior, hands-on AI engineering position responsible for the architecture, performance, and reliability of AI systems, focusing on designing and shipping production-grade RAG and agentic systems that operate at scale.
Computer Software
Responsibilities
Design, build, and scale the end-to-end AI architecture powering DASH
Own core systems including RAG pipelines, model routing, vector databases, and evaluation infrastructure
Ship production-grade LLM systems that handle real-world complexity and scale
Optimize RAG pipelines across large, messy datasets (structured financial data + unstructured legal, lease, and broker documents)
Design and implement agentic and multi-agent workflows to automate complex co-broker tasks
Partner closely with evaluation engineers to diagnose hallucinations, drift, and quality issues
Implement evaluation pipelines, safety mechanisms, and architectural fixes to improve output reliability
Optimize AI systems for performance, cost, and scalability (AI FinOps)
Make smart tradeoffs around model choice, token usage, and routing as the platform grows
Qualification
Required
Proven production AI experience - You've designed, built, and shipped real AI / LLM systems — not just prototypes
Deep RAG expertise - Expert-level understanding of Retrieval-Augmented Generation, vector databases, and hybrid semantic search
Strong data pipeline ownership - Experience building ingestion and processing pipelines for both structured (SQL, CSV) and unstructured (PDFs, docs, emails) data
LLM platform fluency - Deep familiarity with OpenAI APIs and experience with other major model providers (Anthropic, Google)
Embeddings + retrieval judgment - Strong intuition for embedding strategies and retrieval techniques, and knowing when to use which
Cloud-native AI systems - Experience deploying and operating AI services on Azure
Evaluation-first mindset - Fluent in evaluation methodologies, hallucination mitigation, drift detection, and quality control — and able to partner deeply with eval teams
Advanced agentic experience - Hands-on experience (or deep applied knowledge) building agentic and multi-agent systems in production
AI FinOps awareness - Understanding how to design AI systems that scale profitably by managing cost, latency, and throughput
Full-time, US-based role
Ability to work PST ± 4 hours
Excellent English communication
Professional references required
Company
Duxre
There’s a fundamental difference between a platform that simply hosts listings and one that helps you run and grow your commercial real estate business.
Funding
Current Stage
Early StageCompany data provided by crunchbase