AI/ML Engineer — LLM Ops & Grounding for Real Estate

Infinityy; New York (NY, CA); Hybrid; USD 120,000 - 210,000; Full time

Benefits offered by this job

Benefits details shared during the 인터뷰

Job summary

Infinityy is building Fin, the AI guide for real estate listings and buyer Q&A, grounding its agents in MLS data and partner sources. You will own prompts, measure quality, and improve agent behavior across narratives, captions, and lead summaries for a multilingual audience.

You will design retrieval pipelines, build evaluation datasets, and work with backend engineers on integration in a C#/.NET platform, ensuring guardrails and fair housing compliance.

Qualifications

• 3+ years in a technical role (AI/ML engineering, data, software engineering or similar), with hands-on LLM work in production.
• Strong experience with prompt engineering and agent configuration (OpenAI, Anthropic, LangChain, LlamaIndex, Semantic Kernel or similar).
• Python or TypeScript for scripting, data processing and experiment automation.
• Working knowledge of vector databases, embeddings and retrieval.
• Experience with LLMOps and eval tools (LangSmith, Weights & Biases, OpenAI Evals or similar).
• An analytical, experiment-driven approach; you can explain why a model behaves the way it does.
• Clear writing for documenting prompts, experiments and reproducible workflows.

Responsibilities

• Design, test and improve prompts and agent behavior for Fin across listing narratives, captions, neighborhood insights, buyer Q&A and lead summaries
• Build and maintain retrieval and context pipelines that ground Fin in MLS data, agent-uploaded files and partner data such as Local Logic
• Create evaluation datasets, scoring rubrics and automated evals, including fair housing compliance, accuracy and multilingual quality checks
• Run structured experiments and A/B tests; track model quality, latency and cost in dashboards the team can read
• Work with backend engineers on agent orchestration, tool calls and model integration in our C#/.NET platform
• Monitor production outputs for errors, bias and drift; set up guardrails and fixes
• Help shape new AI features with product and design, such as letting agents train Fin on extra context
• Document prompt patterns, eval methods and best practices so the whole team can build on them

Skills

prompt engineering

agent configuration

Python

TypeScript

vector databases

embeddings

retrieval

LLMOps

eval tools

analytical mindset

Tools

OpenAI

Anthropic

LangChain

LlamaIndex

Semantic Kernel

LangSmith

Weights & Biases

OpenAI Evals

Job description

Infinityy is building Fin, the AI guide for real estate listings and buyer Q&A, grounding its agents in MLS data and partner sources. You will own prompts, measure quality, and improve agent behavior across narratives, captions, and lead summaries for a multilingual audience.

You will design retrieval pipelines, build evaluation datasets, and work with backend engineers on integration in a C#/.NET platform, ensuring guardrails and fair housing compliance.