Plaud Senior AI Engineer designs and builds AI services for converting recordings into summaries and notes, focusing on LLM orchestration and model routing. Responsible for scaling AI generation systems and improving AI output quality.
Responsibilities
Design and build the AI services that turn recordings into summaries and notes, from LLM orchestration and model routing to delivering results reliably to millions of users.
Build generation pipelines that adapt to each recording's content, context and user preferences, choosing the right models, strategies and level of detail for each one.
Solve hard scaling problems: keep AI generation highly available across multiple model providers, absorb large and spiky workloads, and keep results fast at any size.
Turn real-world usage and evaluation signals into continuous, measurable quality gains, so that model, prompt and pipeline changes improve the product release after release without regressions.
Partner with AI research and evaluation teams to take new models and capabilities from experiment to production, and use AI tooling to speed up your own development.
Qualification
Proven experience scaling high-volume
Required
5+ years of backend engineering experience, with strong production skills in Python or Go.
Hands-on experience building and running LLM- or ML-powered services in production, including model integration and graceful fallback across providers. Prototypes and internal tooling alone do not qualify.
Proven experience scaling high-volume, asynchronous distributed systems for availability and performance.
Experience measuring and improving AI output quality with data, e.g. automated evaluation, judges calibrated against human ratings, or controlled experiments.
Strong judgment on tradeoffs between AI quality, latency and cost.
Preferred
Experience building systems that systematically optimize prompts, pipelines or model choices from evaluation and production feedback, e.g. programmatic prompt optimization or reflective optimizers.
Experience with speech or transcription systems (ASR, speaker identification, audio processing).
Experience serving and scaling models on GPUs.
Experience with personalization or adaptive generation driven by user context.
Experience with cloud-native infrastructure and multi-region deployment.