Research Engineer - Geo-Distributed Inference

Pluralis Research
United States, Australia
Workplace: RemoteFull timeFunction: Research & Scientific (R&D)Skills: ["Mission alignment","Hands-on execution","Research communication","Problem-solving","Systems thinking"]

Build and own the geo-distributed inference stack that powers permissionless Protocol Learning. You’ll design end-to-end inference pipeline execution (pipeline-parallel scheduling, placement/routing, transport, serving engine, and failure handling) and invent new methods to run fast inference on consumer hardware over the public internet with nodes joining/leaving and weights changing mid-run. You’ll also evolve the rollout pipeline into a reliable serving layer for trained models used for RL training.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Pluralis Research
Pluralis Research
5 days ago

Research Engineer - Geo-Distributed Inference

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 10 hours agoStatus: Live

Job Summary

Build and own the geo-distributed inference stack that powers permissionless Protocol Learning. You’ll design end-to-end inference pipeline execution (pipeline-parallel scheduling, placement/routing, transport, serving engine, and failure handling) and invent new methods to run fast inference on consumer hardware over the public internet with nodes joining/leaving and weights changing mid-run. You’ll also evolve the rollout pipeline into a reliable serving layer for trained models used for RL training.
Location: United States, Australia
Workplace: Remote
Employment Type: Full time
Job Function: Research & Scientific (R&D)

Key Responsibilities

  • •Own the inference stack end-to-end, including pipeline-parallel execution, placement/routing, transport, serving engine, and failure handling.
  • •Design new algorithms to make inference fast on consumer hardware over public internet conditions, validating them and moving them to production.
  • •Run the rollout pipeline for reinforcement learning training today and convert it into the serving layer after models are trained.
  • •Build systems that remain fast and reliable in a permissionless setting where nodes join/leave mid-run and weights can change under the server.
  • •Continuously improve throughput and reliability of the geo-distributed inference pipeline under real-world constraints.

Pay and Benefits

Perks:EquityRemote WorkRelocationVisa Sponsorship

Key Requirements

  • •Shipped serving systems: shipped serving-engine internals or built a large-scale inference system hands-on.
  • •Research ability shown via publications (papers/blogposts) in distributed inference or nearby fields, or unpublished work you can walk through.
  • •Experience with low-bandwidth, high-latency systems similar to public-internet conditions.
  • •You can keep inference fast and reliable under changing conditions like nodes joining/leaving mid-run and weights updating.
  • •Mission alignment with Protocol Learning for collective, trustless, and sovereign AI.
Experience:Machine learningDistributed systemsLLM servingReinforcement learningDecentralized trainingConsumer hardware
Skills:Mission alignmentHands-on executionResearch communicationProblem-solvingSystems thinking
Languages:English
Tech Stack:Distributed inferenceLLM serving systemsPipeline parallelismLow-bandwidth networkingHigh-latency networksReinforcement learningRL post-trainingApple siliconMLXP2P networkingNAT traversalOpen-weight AIOpen-source AI

Eligibility

Work Authorization:Sponsorship available.

Company Brief

Pluralis Research
Develops Protocol Learning for decentralized, multi‑participant training of foundation models so models remain unmaterialized and community‑owned, enabling open-source large‑scale AI without single‑party control.
Industry: AI & Machine Learning
Company Size: Micro (1 to 10 employees)
Growth: Early Stage Startup
Funding: Seed
Founded: 2024
Glassdoor
Glassdoor: 4.0
WebsiteLinkedIn