Where Neuroscience Meets Agentic AI

About Raynmaker

We're building RaynBrain , the first agentic AI platform for complex conversations.

Grounded in machine learning, neuroscience, and forensic linguistics, RaynBrain powers autonomous systems that interpret, adapt, and act in real time. These systems turn raw leads into revenue without scripts, static flows, or human handoffs.

Enterprise power without the bloat. Raynmaker helps small teams move faster, convert more, and never waste another lead. We replace the complexity of traditional sales stacks with AI that listens, reasons, and closes.

The Role

We're hiring a Senior AI/ML Engineer to architect and scale the core intelligence behind our platform. This role spans systems design, ML engineering, and LLM integration. It sits at the intersection of infrastructure and applied AI.

You will design, build, and optimize the pipelines and agent systems that drive live customer interactions. That includes retrieval-augmented generation (RAG), scoring models, vector search, real-time streaming inference, memory management, and reinforcement learning systems. All of it is deployed in production and built to scale.

You will partner with engineering leadership to take ideas from whiteboard to production quickly and own key decisions around performance, cost efficiency, and reliability.

What You'll Build

• RAG pipelines using Milvus, Weaviate, Pinecone, or Zilliz
• Custom LLM deployments with fine-tuning, inference routing, and token optimization
• Tool-calling and agent flows supporting complex, multi-step decisions
• Reinforcement learning systems to evolve agent behavior over time
• Streaming inference pipelines for voice, chat, and other live interactions
• Multi-tenant ML infrastructure with robust data isolation and observability

Core Responsibilities

LLM, Retrieval, and Agent Systems

• Design and optimize production-grade RAG systems
• Build ranking, scoring, and routing models for live inference
• Architect tool-calling flows, agent memory, and multi-turn reasoning
• Optimize token usage, caching, and cost-performance tradeoffs
• Maintain and enrich vector knowledge bases

ML Engineering and Data Infrastructure

• Build real-time and batch pipelines for ingestion, training, and inference
• Deploy and monitor reinforcement learning systems
• Own the ML model lifecycle across development, evaluation, deployment, and tuning
• Drive continuous optimization across latency, cost, and performance

Systems Integration and Deployment

• Build and maintain ML APIs and microservices using Docker and Kubernetes
• Support streaming interaction layers including voice and WebSockets
• Ensure production reliability, monitoring, and scale
• Collaborate cross-functionally on platform-wide architecture and data contracts

You Should Have

• 7+ years of experience in ML, AI, or data engineering roles
• Expert-level Python for backend, ML workflows, and orchestration
• Experience with modern LLM frameworks such as LangChain or LangGraph
• Deep knowledge of vector databases and retrieval systems
• Production experience with reinforcement learning
• Comfort with distributed systems, Docker, and Kubernetes
• Experience building and maintaining streaming or real-time pipelines
• A track record of shipping complex systems that work in production

Nice to Have

• Familiarity with AWS ML stack including SageMaker or Bedrock
• Experience with Kafka, Kinesis, or Pulsar
• Knowledge of model compression, quantization, or accelerated inference
• CRM or sales tech background such as Salesforce or HubSpot

Why Raynmaker

• High Impact : We are building for the 99 percent of businesses left behind by legacy software. Your work will help small teams win with tech that is fast, affordable, and deeply capable.
• Hard Problems : We are solving real-time inference, agent coordination, and scalable autonomy, not just wrapping APIs.
• Applied Intelligence : We combine machine learning with neuroscience and forensic linguistics to model not just what people say but how and why they say it. You'll build agents that detect hesitation patterns, sentiment shifts, and objection timing - then adapt strategy in real time based on behavioral cues, not just keywords.
• Deep Ownership : You will shape architecture and systems from end to end, not just optimize what someone else scoped.

This isn't research for research's sake. This is production-grade intelligence solving real problems for real businesses, every single day. If that's the kind of impact you want, we'd love to meet you.