Lead Architect _ Intelligent Voice Assistant Systems

Raya Workforce

Milford, MA, USA

Published: 6/14/2022

Technology

Full Time

Job Description

Location: Flexible | Remote-First
Role Type: Full-Time | Founding Team Position

Are you excited about creating the next wave of intelligent, voice-enabled systems? We’re looking for a forward-thinking technologist to help define the architecture of an advanced Virtual Personal Assistant (VPA) platform—one built where natural conversation, on-device performance, and cutting-edge AI meet.

What You’ll Be Building

You’ll guide the full architectural vision for a modern VPA—think SIRI’s concept, but rebuilt with today’s most powerful technologies. Expect to work on a blend of local and cloud inference, high-accuracy speech recognition systems, and deep integration across Apple hardware and software.

Core Tech Stack

Devices & Platforms: iPhone, iPad, Mac Mini M4 Pro

Key Technologies & Frameworks:

Apple Development Tools (Swift, Core ML)
LangGraph for agent orchestration
Large Language Models (LLMs)
Retrieval-Augmented Generation (RAG)
ASR frameworks: OpenAI Whisper, Meta Wave2Vec-BERT
Hugging Face exporters and Swift-based transformer tooling
Accent-adaptive tuning and speech personalization
PyTorch, JAX, Ray
Distributed data pipelines with Beam, Spark, Dask, and Ray

Your Responsibilities

Architect and enhance a hybrid cloud/on-device system optimized for real-time assistant performance
Develop scalable, privacy-centric ASR and LLM workflows for natural language understanding
Tune transformer models for Apple-native deployment using Core ML and associated tooling
Drive inclusive, multilingual, and accent-resilient voice experiences
Work closely with a small, passionate team of engineers, designers, and product thinkers

Ideal Candidate

You might thrive here if you:

Have experience designing large, complex AI systems and delivering production-level voice interfaces
Bring 3+ years of hands-on work in AI/ML, particularly in Automatic Speech Recognition
Are comfortable with open-source ML ecosystems like PyTorch, JAX, and Ray
Understand distributed processing tools such as Beam, Spark, Dask, or Ray
Are proficient in multiple programming languages (C++, Python, Java preferred)
Enjoy experimenting—whether quickly prototyping or drafting detailed architectural plans
Care about accessibility, user experience, and creating fair voice interactions across accents and languages

Why You’ll Love Working With Us

This isn’t just another software project—it’s a chance to redefine what personalized, voice-driven AI can become. As a founding member, you’ll help build a system that listens, adapts, and responds in the most human-centered way possible.

And here’s a unique perk: you’ll get the opportunity to travel to the Eastern Caribbean—on us—to field-test the product in a beautiful, tropical environment. Innovation meets island energy.