Senior AI Engineer – Neural Networks, Voice & Autonomous Agents

About SmartAdvocate & SmartIntelligence

SmartAdvocate is a leading legal case management software engineered to streamline operations, litigation workflows, and client management for law firms. Through our proprietary AI initiative, SmartIntelligence, we are building next-generation artificial intelligence tools directly into our platform.

We are expanding our engineering team to develop state-of-the-art voice agents for automated intake and autonomous background agents for complex legal document processing, search, and workflow automation.

Position overview

We are seeking a Senior AI Engineer specializing in Neural Networks, Voice, and Autonomous Background Agents to join our team. In this role, you will move beyond standard API integrations to design, fine-tune, and deploy proprietary deep learning models tailored to legal and medical domain requirements.

You will lead the development of ultra-low latency voice agents capable of managing complex phone intakes while accurately handling specialized legal and medical terminology. Additionally, you will build and optimize fine-tuned Small Language Models (SLMs) and custom vector retrieval architectures to power autonomous background agents across the SmartAdvocate platform.

Key Responsibilities

Voice Agent & Speech Architecture

• Fine-tune and optimize Speech-to-Text (STT) and Text-to-Speech (TTS) neural models on domain-specific datasets such as legal terminology, medical ICD codes, and mass tort nomenclature.
• Engineer end-to-end streaming audio pipelines using WebRTC, voice activity detection (VAD), and low-latency audio processing to enable natural, real-time conversational flows.
• Reduce voice response latency to sub-second levels for automated client intake and phone workflows.

Background Agents & Small Language Models (SLMs)

• Train, fine-tune, and quantize open-weight models such as Llama, Mistral, and Qwen for local and private cloud deployments to minimize operational inference costs.
• Build multi-step autonomous background agents for automated case file classification, medical record summarization, and task routing.
• Implement robust agent monitoring, logit analysis, and debugging frameworks to diagnose attention drift and prevent hallucinations in automated workflows.

Information Retrieval & Neural Search

• Design custom dense retrieval models and neural re-rankers optimized for long-form court filings, legal briefs, and client records.
• Enhance Retrieval-Augmented Generation (RAG) pipelines within SmartIntelligence to ensure high precision and zero-hallucination data extraction.

Qualifications & Requirements

Technical Core

• Deep Learning Expertise: Strong foundation in deep neural network architectures, attention mechanisms, Transformer models, and gradient optimization strategies.
• Framework Proficiency: Hands-on mastery of PyTorch, Hugging Face Transformers, ONNX Runtime, TensorRT, and vLLM or similar high-throughput inference engines.
• Voice & Audio AI: Proven experience with speech recognition models such as Whisper or Wav2Vec2, streaming TTS architectures, and raw audio tensor manipulation.
• Agentic Workflows & RAG: Track record of deploying production-grade agent architectures, custom vector search spaces, and specialized embedding models.
• Engineering Stack: Advanced knowledge of Python, asynchronous programming via asyncio, gRPC/REST APIs, Docker, and GPU-accelerated cloud infrastructure across AWS, Azure, or GCP.

Legal AI Background (Highly Preferred)

• Legal AI Industry & Law Firm Experience: Prior experience developing artificial intelligence applications directly within a law firm environment, or at a legal technology entity that builds AI software actively utilized by law firms, is highly preferred.
• Domain Adaptation: Proven success adapting deep learning architectures, LLMs, and speech models to handle legal workflows, statutory logic, complex court filings, or medical-legal documentation.

What We Offer

• The opportunity to shape core intellectual property in the legal technology industry.
• Direct involvement in scaling SmartIntelligence across hundreds of law firms nationwide.
• Collaborative engineering environment focused on privacy, proprietary capability, and technical excellence.

How we hire

We read work, not keywords. Our process is deliberately short:

• Resume and portfolio review
• A 45-minute conversation about systems y…