About Us We're a venture-backed startup building an innovative AI-driven platform called Proxi for policy professionals. Our product helps users navigate complex regulatory, legislative and emerging compliance environments through a voice-driven interface powered by our cutting-edge semantic search pipeline operating over our immense legislative and regulatory data corpus. Our team's experience spans roles across the White House, State Department, K&S Law, Microsoft and Datadog.
The Role You'll join our founding engineering team to enhance and scale the AI foundation that powers our competitive advantage. Working with our existing ML systems, you'll help expand our capabilities in natural language understanding, document processing, and personalized relevance with a particular focus on creating exceptional voice-driven AI interactions.
Location: New York City
Key Responsibilities
Expand our learning-to-rank system and integrated feedback mechanisms
Enhance entity recognition and classification systems for legislative texts
Optimize multi-model embedding architectures for improved semantic understanding
Extend our document enrichment pipeline for new legislative and regulatory formats
Build and refine natural voice interactions that feel human-like and intuitive
Design and implement context-aware voice agents that maintain conversational coherence
Create voice-optimized response generation systems that communicate complex information clearly
Refine personalization systems that enable our platform to learn from user interactions
Collaborate with the engineering team on AI strategy implementation
Requirements
Advanced knowledge of Python for ML development and production systems
Strong experience with LLM orchestration and prompt engineering
Experience building stateful AI agents and multi-agent systems
Practical experience with leading LLM providers and their APIs
Strong background in NLP, particularly entity recognition and extraction
Experience with conversational AI systems and natural dialogue flows
Knowledge of voice processing technologies and speech synthesis optimization
Experience designing voice interfaces that handle complex, domain-specific vocabularies
Experience with vector embeddings and semantic search implementations
Familiarity with SQL databases and vector storage solutions
Knowledge of asynchronous programming patterns in Python
Experience with API development and data validation
Understanding of cloud infrastructure for ML deployments
Nice to Have
Experience with legislative, regulatory, compliance or legal text processing
Background in learning-to-rank systems or relevance optimization
Knowledge of modern search technologies
Experience with real-time data processing pipelines
Background in contextual search or semantic search systems
Knowledge of voice streaming architectures and low-latency response systems
$150,000 - $190,000 a year
Proxi is an AI-native platform that breaks away from the 'copilot' model where LLMs are merely added as features to existing technology. We act as an extension of the user and their interests, working 24/7 to perform work that would otherwise take hours.
We're not looking for passengers; we want driven innovators with a passion for NLP and a hunger to build from the ground up. If you're obsessed with pushing the boundaries of natural language understanding, document processing, and personalized relevance, and you thrive in a fast-paced, high-stakes startup environment, this is your moment. We're building a category-defining product - don't just apply; make your mark.