People

Applied Research Engineer — Memory Architecture & Entity Coherence

US · Remote · Senior / Staff (6+ years)

The research agenda is about making memory more accurate, more scalable, and capable of things it cannot do today — from experiment to deployed system.

What you will work on

Memory architecture

How memory should be structured as enterprise knowledge types and volumes grow.

Entity Coherence at scale

Extend accuracy and handle new failure modes at enterprise scale.

Retrieval quality and evaluation

Ground truth datasets and metrics for temporally sensitive correctness.

Model fine-tuning

Dataset creation, training, and evaluation for extraction and ranking in the memory domain.

What we are looking for

Required

  • 6+ years production engineering with backend systems depth
  • Research methodology and hands-on fine-tuning (PyTorch)
  • Retrieval systems in production
  • Python plus Rust, Go, or C++

Strongly preferred

  • Published or open-source work in retrieval, memory, or entity resolution
  • Knowledge graph construction at scale

Not a fit if

  • Model research without shipping to production