GRPO Jobs - Remote & On-site

GRPO jobs hiring now at vetted startups and product companies on Alion, updated daily. Average salary around $313,042/year. Compare remote, hybrid and on-site positions with relocation and visa sponsorship. Browse and apply today.

AI/ML
Data Science
Backend
Frontend
Mobile
DevOps
Web3
Games
Hardware
Robotics
Security
QA
Leadership
Networking
Product
Design
Analytics
Support
Solutions
Enterprise Apps
Quantum

Kakao

kakaocorp.com
kakaocorp.com • HQ: Jeju, South Korea • Ride Hailing • Digital Banks & Neobanks • Payment Processing & Gateways • Consumer Internet Groups • Messaging Apps • 1001-5000 employees • Est. 1995 • Trust C • Hiring 23
HQ: Jeju, South Korea • Ride Hailing • Digital Banks & Neobanks • Payment Processing & Gateways • Consumer Internet Groups • Messaging Apps • 1001-5000 employees • Est. 1995 • Trust C • Hiring 23
Kakao is a South Korean internet company whose listed entity traces back to Daum Communications, founded in 1995, and which took its present form when Daum merged with the messaging startup Kakao in 2014. KakaoTalk, launched in 2010, is used by almost the entire South Korean population and functions as the country's default communication layer, carrying payments, shopping, gifting and content on top of messaging. Around it the group has built Kakao Pay, Kakao Bank, the Kakao T mobility app, Kakao Entertainment for webtoons and music, and the Kanana family of AI assistants.
Shin-a Chung · Chief Executive Officer · decision-maker
Beom-su Kim · Founder · founder
Careers mailbox is known
Verified live · 2 hours ago 2 months ago Long shot · 6%

AI Research Engineer (Search & Agent) (경력)

In office (Seongnam, South Korea) • Full-Time
C#
C#
.NET
AI/ML
Knowledge Distillation
Function Calling
AI Agents
LLM
Reranking
Hybrid Search
Contrastive Learning
PPO
GRPO
arXiv
Tool Use
Model Distillation
+6
Apply

Cantina

cantina.com
cantina.com • HQ: San Francisco, United States • Social Networks • Multimodal AI • Generative Video • LLM & Generative AI • 51-200 employees • Est. 2023 • Startup • Trust A • Hiring 19
HQ: San Francisco, United States • Social Networks • Multimodal AI • Generative Video • LLM & Generative AI • 51-200 employees • Est. 2023 • Startup • Trust A • Hiring 19
Cantina is a social artificial intelligence technology company based in San Francisco, California, and founded in 2023. The company provides a platform where users can create, interact with, and share multimodal AI characters that feature unique personalities and the ability to generate video content. It operates primarily through a mobile application and web interface, focusing on the intersection of generative AI and social media for a global audience of creators and consumers.
No named decision-maker yet.
No careers mailbox known
Verified live · 2 hours ago 2 months ago Long shot · 24%

Machine Learning Engineer - Voice Conversion

$200k – $220k per year • In office • Full-Time
C++
C++
PyTorch C++
AI/ML
DeepSpeed
CUDA Toolkit
Multimodal AI
Knowledge Distillation
Speech Recognition
PyTorch
Tokenization
Synthetic Data
CUDA
Triton
DPO
GRPO
FSDP
Text-to-Speech
LLM Guardrails
Model Distillation
+10
Apply
Verified live · 2 hours ago 6 months ago Long shot · 10%

Machine Learning Engineer, TTS

$200k – $220k per year • In office • Full-Time
C++
C++
PyTorch C++
AI/ML
CUDA Toolkit
Knowledge Distillation
Diffusion Models
Speech Recognition
PyTorch
Synthetic Data
CUDA
Triton
DPO
GRPO
Text-to-Speech
LLM Guardrails
Model Distillation
+7
Apply

Together AI

together.ai
together.ai • HQ: San Francisco, United States • Foundation Models • Cloud Platforms (IaaS & PaaS) • AI Compute & Inference • 501-1000 employees • Est. 2022 • Startup • Trust A • Hiring 77, 8 new
HQ: San Francisco, United States • Foundation Models • Cloud Platforms (IaaS & PaaS) • AI Compute & Inference • 501-1000 employees • Est. 2022 • Startup • Trust A • Hiring 77, 8 new
Together AI (Together Computer, Inc.) is a full-stack AI infrastructure and cloud platform headquartered in San Francisco, California. Founded in 2022 by prominent AI researchers and system engineers - including CEO Vipul Ved Prakash, CTO Ce Zhang, Chief Scientist Tri Dao (co-creator of FlashAttention), Chris Ré, and Percy Liang - the company operates as an "AI Native Cloud" designed to train, fine-tune, and deploy open-source generative AI models at scale with high performance and optimized unit economics.
No named decision-maker yet.
No careers mailbox known
Verified live · 1 hour ago 2 months ago Fading · 39%

Forward Deployed Engineer (Inference & Post-Training) - Mandarin Speaking

≈ $73k – $185k per year (Estimated) • Senior • 5+ years exp • Singapore • Remote (Singapore) • Senior
Python
AI/ML
Cursor
LoRA
vLLM
ElevenLabs
Fine-tuning
RLHF
Reinforcement Learning
Quantization
SGLang
TensorRT-LLM
LLM
Together AI
DPO
SFT
GRPO
Post-training
Pre-training
Speculative Decoding
KV Cache
TensorRT
PEFT
+14
Apply
Verified live · 1 hour ago Visa sponsor 5 months ago Long shot · 18%

Forward Deployed Engineer (Inference & Post-Training)

≈ $176k – $331k per year (Estimated) • Senior • 5+ years exp • San Francisco • Remote (United States) • Full-Time • Senior
Python
AI/ML
Cursor
LoRA
vLLM
ElevenLabs
Fine-tuning
RLHF
Reinforcement Learning
Quantization
SGLang
TensorRT-LLM
LLM
Together AI
DPO
SFT
GRPO
Post-training
Pre-training
Speculative Decoding
KV Cache
TensorRT
PEFT
+14
Apply

JPMorganChase

jpmorganchase.com
jpmorganchase.com • HQ: New York, United States • Wealth Management & Financial Advisors • Asset Management & Funds • Investment Banking & M&A Advisory • Commercial & Retail Banks • 5000+ employees • Est. 1799 • Trust A • Hiring 2895, 303 new
HQ: New York, United States • Wealth Management & Financial Advisors • Asset Management & Funds • Investment Banking & M&A Advisory • Commercial & Retail Banks • 5000+ employees • Est. 1799 • Trust A • Hiring 2895, 303 new
JPMorganChase is the largest bank in the United States by assets and one of the most systemically important financial institutions in the world, with a lineage running back through more than a thousand predecessor firms to the 1799 founding of the Bank of the Manhattan Company. It combines a dominant investment bank and markets business with Chase, the largest retail banking franchise in America, plus commercial banking and asset and wealth management. Headquartered in New York, the group is unusual among banks for the scale of its technology spending, running one of the largest engineering organisations of any financial institution and deploying its own internal AI platform across the firm.
Richard Dulik · Executive Officer · decision-maker
Thomas Chase · Executive Officer, Director · decision-maker
Marianne Lake · Chief Executive Officer, Consumer and Community Banking · decision-maker
Jeremy Barnum · Chief Financial Officer · decision-maker
No careers mailbox known
Verified live · 1 hour ago Visa sponsor 2 months ago Long shot · 11%

Executive Director Machine Learning Engineer-MLOps

≈ $254k – $545k per year (Estimated) • Head+ • 10+ years exp • In office (Palo Alto, United States) • Bachelor's Degree • Architect
Python
AI/ML
vLLM
CUDA Toolkit
Fine-tuning
Reinforcement Learning
Quantization
AWQ
Transformers
Ray
CUDA
GRPO
Machine Learning
DevOps
AWS
Kubernetes
Amazon ECS
+7
Apply

Boundless

boundless.network
boundless.network • AI Agents • AI Compute & Inference • Hiring 4
AI Agents • AI Compute & Inference • Hiring 4
Boundless is the inference partner that helps AI-native companies scale their AI usage, lower costs, and keep quality high.
No named decision-maker yet.
No careers mailbox known
Top 25% payVerified live · 4 hours ago 2 months ago Long shot · 11%

Applied AI/ML Engineer

$175k – $250k per year • Remote (United States) • Full-Time
Python
AI/ML
vLLM
CUDA Toolkit
Reinforcement Learning
Quantization
SGLang
TensorRT-LLM
PyTorch
LLM
CUDA
DPO
SFT
PPO
GRPO
Post-training
Megatron-LM
LLM Evaluation
Speculative Decoding
KV Cache
TensorRT
SkyPilot
Ray
FSDP
DevOps
SLURM
Kubernetes
GitHub
+18
Apply

Labelbox

labelbox.com
labelbox.com • HQ: San Francisco, United States • AI Training Data & Annotation • 201-500 employees • Est. 2018 • Startup • Trust A • Hiring 7
HQ: San Francisco, United States • AI Training Data & Annotation • 201-500 employees • Est. 2018 • Startup • Trust A • Hiring 7
Labelbox is a San Francisco company founded in 2018 that provides a data engine for training and evaluating machine learning models. Its platform combines annotation tooling, model-assisted labelling and quality analytics, and it operates an expert workforce marketplace for tasks that require domain knowledge. Customers range from computer vision teams to frontier labs sourcing human preference data.
No named decision-maker yet.
No careers mailbox known
Verified live · 3 hours ago Visa sponsor 3 months ago Long shot · 24%

Staff ML Engineer, Agent Training & Environments

$250k – $280k per year • Equity • Staff+ • 3+ years exp • In office (San Francisco, United States) • Staff
Python
JavaScript
Java
Kotlin
TypeScript
Node JS
Databases
MySQL
PostgreSQL
Databricks
Apache Kafka
Google Cloud Spanner
AI/ML
Fine-tuning
AI Agents
LLM
DPO
SFT
PPO
GRPO
Post-training
Frontend
Redux
GraphQL
React.js
DevOps
GCP
Kubernetes
Analytics
Azure Data Factory
+17
Apply

INTSIG

intsig.com
intsig.com • HQ: Shanghai, China • B2B Company & Contact Data • Consumer & Utility Software • Document AI • Document Management • 1001-5000 employees • Est. 2006 • Trust B • Hiring 77, 1 new
HQ: Shanghai, China • B2B Company & Contact Data • Consumer & Utility Software • Document AI • Document Management • 1001-5000 employees • Est. 2006 • Trust B • Hiring 77, 1 new
INTSIG is a Chinese artificial intelligence company headquartered in Shanghai that develops text intelligence products, including the CamScanner document scanning app and the CamCard business card app for consumers and the TextIn and Qixin Insight suites for enterprise document processing and commercial data. Founded in 2006 by a group of scientists, it reports users in some 200 countries and regions, about 199 million monthly active users and enterprise clients in nearly 30 industries, and it went public in 2024. Hiring centres on multimodal and large model researchers, image and document recognition engineers, AI product managers and interns.
No named decision-maker yet.
No careers mailbox known
Verified live · 1 hour ago 3 months ago Long shot · 5%

27届校招-数据算法工程师(J14427)

In office (Suzhou, China)
Python
Databases
Milvus
FAISS
AI/ML
vLLM
DPO
PPO
GRPO
Megatron-LM
GraphRAG
KV Cache
+2
Apply

Waymo

waymo.com
waymo.com • HQ: Mountain View, United States • Ride Hailing • Smart Mobility • Autonomous Driving • 5000+ employees • Est. 2009 • Trust B • Hiring 356, 21 new
HQ: Mountain View, United States • Ride Hailing • Smart Mobility • Autonomous Driving • 5000+ employees • Est. 2009 • Trust B • Hiring 356, 21 new
Waymo is an American autonomous driving company that began as the Google self-driving car project in 2009 and became a standalone Alphabet subsidiary in 2016. It operates the largest fully driverless commercial ride-hailing service in the world through the Waymo One app, carrying paying passengers with no human in the vehicle across Phoenix, San Francisco, Los Angeles, Austin and an expanding list of cities. Headquartered in Mountain View, California, the company builds its own sensor suite and the Waymo Driver software stack, integrates them into vehicles from partners such as Jaguar, Zeekr and Hyundai, and also develops autonomous trucking technology.
Dmitri Dolgov · Co-Chief Executive Officer & Chief Technology Officer · decision-maker
Tekedra Mawakana · Co-Chief Executive Officer · decision-maker
Sebastian Thrun · Founder of the Google Self-Driving Car Project · founder
Anthony Levandowski · Founding Engineer
No careers mailbox known
Top 25% payVerified live · 1 hour ago Visa sponsor 3 months ago Fading · 35%

Staff Machine Learning Engineer - Vision-Language Foundation Models

$251k – $310k per year • Equity • Staff+ • 8+ years exp • Hybrid (Mountain View, United States) • Master's Degree • Full-Time • Staff
Python
AI/ML
Fine-tuning
RLHF
Reinforcement Learning
JAX
Multimodal AI
Chain-of-Thought
VLM
PyTorch
Gemini
Tokenization
Synthetic Data
DPO
SFT
PPO
GRPO
Post-training
Pre-training
Megatron-LM
FSDP
TPU
RLAIF
Robotics
Reinforcement Learning
+15
Apply

Nebius

nebius.com
nebius.com • HQ: Amsterdam, Netherlands • Data Centers & Colocation • Cloud Platforms (IaaS & PaaS) • AI Compute & Inference • 1001-5000 employees • Trust B • Hiring 335, 32 new
HQ: Amsterdam, Netherlands • Data Centers & Colocation • Cloud Platforms (IaaS & PaaS) • AI Compute & Inference • 1001-5000 employees • Trust B • Hiring 335, 32 new
Nebius is an international technology company that specializes in building full-stack artificial intelligence infrastructure and high-performance cloud GPU platforms for AI model development. Headquartered in Amsterdam, Netherlands, the firm operates energy-efficient data centers across Europe and North America to provide scalable compute, storage, and software tools for machine learning workloads.
No named decision-maker yet.
Careers mailbox is known
Top 25% payVerified live · 3 hours ago 3 months ago Long shot · 16%

Senior Machine Learning Engineer, Model Training and Reinforcement Learning

$195k – $262k per year • Senior • Palo Alto • Remote (United States) • Senior
Python
AI/ML
DeepSpeed
RLHF
Function Calling
PyTorch
LLM
Ray
Synthetic Data
DPO
SFT
PPO
GRPO
Post-training
Megatron-LM
FSDP
Agentic Workflows
Tool Use
RLAIF
Machine Learning
CUDA Toolkit
TRL
Transformers
CUDA
Triton
NCCL
InfiniBand
Reward Modeling
DevOps
SLURM
Kubernetes
+21
Apply

SenseTime

sensetime.com
sensetime.com • HQ: Shanghai, China • LLM & Generative AI • Foundation Models • Computer Vision • 1001-5000 employees • Est. 2014 • Trust C • Hiring 67, 4 new
HQ: Shanghai, China • LLM & Generative AI • Foundation Models • Computer Vision • 1001-5000 employees • Est. 2014 • Trust C • Hiring 67, 4 new
SenseTime is a Chinese artificial intelligence company founded in 2014 out of computer vision research at the Chinese University of Hong Kong, and one of the country's largest AI software vendors. It built its business on face and image recognition sold to city governments, banks, retailers and phone makers, running on SenseCore, its own large-scale training and inference infrastructure. Since 2023 the company has pivoted toward generative AI with the SenseNova model family and the SenseChat assistant, while also selling autonomous driving software, and it has operated under United States sanctions since 2021.
Wang Xiaogang · Co-Founder & Chief Scientist · decision-maker
Xu Li · Co-Founder, Chairman & Chief Executive Officer · decision-maker
Tang Xiao'ou · Co-Founder · founder
Careers mailbox is known
Likely inactiveVerified live · 1 hour ago 3 months ago Long shot · 4%

大模型强化学习研究员

In office (Beijing, China) • Full-Time • Chinese: C1
Python
AI/ML
Cursor
Claude Code
AI Agents
PyTorch
PPO
GRPO
DevOps
GitHub
Apply

NIO

nio.com
nio.com • HQ: Shanghai, China • 1001-5000 employees • Est. 2014 • Trust C • Hiring 255, 16 new
HQ: Shanghai, China • 1001-5000 employees • Est. 2014 • Trust C • Hiring 255, 16 new
NIO designs premium electric vehicles with an extensive battery swapping network. Its cars include a proprietary digital assistant and advanced driver assistance system. The company operates member lounges and services across China and parts of Europe.
No named decision-maker yet.
Careers mailbox is known
Likely inactiveVerified live · 1 hour ago 3 months ago Long shot · 4%

智能辅助驾驶端到端算法开发实习生

In office (Shanghai, China) • Internship • Intern • Chinese: C1
Python
AI/ML
VLM
PyTorch
DPO
GRPO
Vision-Language-Action
Apply
Likely inactiveVerified live · 1 hour ago 4 months ago Long shot · 4%

智能辅助驾驶-大模型后训练算法工程师/专家

In office (Beijing, China) • Full-Time • Senior • Chinese: C1
VLM
DPO
SFT
GRPO
Apply
Likely inactiveVerified live · 1 hour ago 4 months ago Long shot · 3%

智能辅助驾驶强化学习算法工程师

In office (Shanghai, China) • Full-Time
RLHF
Transformers
LLM
DPO
PPO
GRPO
World Models
Reward Modeling
Robotics
CARLA
+1
Apply

IFS

ifs.com
ifs.com • HQ: Linköping, Sweden • Manufacturing & Industrial Software • Field Service Management Software • ERP & Business Management Software • 5000+ employees • Est. 1983 • Trust B • Hiring 221, 13 new
HQ: Linköping, Sweden • Manufacturing & Industrial Software • Field Service Management Software • ERP & Business Management Software • 5000+ employees • Est. 1983 • Trust B • Hiring 221, 13 new
IFS is a Swedish enterprise software company founded in Linkoping in 1983 that sells applications for organisations whose value depends on physical assets and field work rather than office processes. Its Cloud product combines enterprise resource planning, asset management and field service management in one model, which suits aerospace and defence, energy and utilities, manufacturing, construction and telecommunications customers who plan maintenance, parts and technicians together. Headquartered in Linkoping and majority owned by the private equity firms EQT and Hg, it has pushed hard into industrial artificial intelligence.
Matthias Heiden · Chief Financial Officer · decision-maker
Mark Moffat · Chief Executive Officer · decision-maker
Careers mailbox is known
Top 25% payVerified live · 9 min ago 3 months ago Long shot · 12%

Data Scientist - Agentic AI Systems - IFS Loops

$140k – $150k per year • Middle • 4+ years exp • Hybrid (Palo Alto, United States) • Master's Degree • Full-Time • Senior • English
Python
SQL
AI/ML
LangGraph
AutoGen
LangChain
Fine-tuning
Embeddings
RLHF
Reinforcement Learning
Scikit-learn
Function Calling
AI Agents
PyTorch
LLM
RAG
DPO
SFT
GRPO
Post-training
Tool Use
+12
Apply

Console

console.com
console.com • HQ: San Francisco, United States • IT Help Desk & Service Desk • AI Agents • IT Service & Asset Management • 11-50 employees • Startup • Trust D • Hiring 15
HQ: San Francisco, United States • IT Help Desk & Service Desk • AI Agents • IT Service & Asset Management • 11-50 employees • Startup • Trust D • Hiring 15
Console is an American software company that builds an AI-native IT service management platform. Its AI agents handle employee IT requests such as access provisioning, password resets and troubleshooting inside chat tools like Slack, and help IT teams automate support workflows instead of working tickets by hand. The company is headquartered in San Francisco, California, United States.
No named decision-maker yet.
Careers mailbox is known
Likely inactiveTop 25% pay 3 months ago Long shot · 6%

Research Engineer

$200k – $350k per year • In office (San Francisco, United States) • Full-Time
Databricks
AI/ML
Cursor
DSPy
LoRA
Fine-tuning
Reinforcement Learning
Knowledge Distillation
AI Agents
PEFT
QLoRA
OpenAI
DPO
SFT
GRPO
Post-training
LLM Evaluation
Model Distillation
Reward Modeling
Machine Learning
DevOps
GitHub
Design
Figma
Management
Slack
Stripe
+15
Apply

AgiBot

agibot.com.cn
agibot.com.cn • HQ: Shanghai, China • Robotic Software & Control Systems • Robotics AI • Humanoid Robots • 1001-5000 employees • Est. 2023 • Trust A • Hiring 436, 2 new
HQ: Shanghai, China • Robotic Software & Control Systems • Robotics AI • Humanoid Robots • 1001-5000 employees • Est. 2023 • Trust A • Hiring 436, 2 new
Headquartered in Shanghai, China, AgiBot - also known as Zhiyuan Robotics - is an innovative artificial intelligence and robotics company specializing in general-purpose humanoid robots and embodied AI systems. The firm develops advanced hardware platforms and specialized operating software that integrate computer vision, motion control, and deep learning capabilities. By producing both bipedal and wheeled robotic systems, it provides scalable automation solutions designed for industrial manufacturing, commercial environments, and research applications.
No named decision-maker yet.
Careers mailbox is known
Verified live · 2 hours ago 3 months ago Likely open · 51%

全模态大模型算法研究员

In office (Beijing, China) • Full-Time • Chinese: C1
Qwen
RLHF
DPO
SFT
PPO
GRPO
GPT-4
Apply
Verified live · 2 hours ago 10 months ago Long shot · 10%

大模型算法工程师-强化学习

In office (Shanghai, China) • Full-Time • Chinese: C1
PPO
GRPO
Apply

Spotify

spotify.com
spotify.com • HQ: Stockholm, Sweden • Advertising • Podcasting • Music Industry • Streaming & OTT Platforms • 5000+ employees • Est. 2006 • Trust B • Hiring 78, 4 new
HQ: Stockholm, Sweden • Advertising • Podcasting • Music Industry • Streaming & OTT Platforms • 5000+ employees • Est. 2006 • Trust B • Hiring 78, 4 new
Spotify is a Swedish audio streaming company founded in Stockholm in 2006 that turned licensed on-demand music into the dominant way people listen. Its catalogue spans roughly a hundred million tracks plus podcasts and audiobooks, delivered through an advertising-supported free tier and a paid Premium subscription to hundreds of millions of monthly users across more than a hundred and eighty markets. The platform pairs personalised recommendation systems such as Discover Weekly with a creator side that gives artists, podcasters and publishers distribution, analytics and monetisation tools.
Daniel Ek · Co-Founder & Executive Chairman · decision-maker
Alex Norstrom · Co-Chief Executive Officer · decision-maker
Gustav Soderstrom · Co-Chief Executive Officer · decision-maker
Martin Lorentzon · Co-Founder · founder
No careers mailbox known
Verified live · 3 hours ago Visa sponsor 3 months ago Long shot · 4%

Research Scientist - Generative Audio

≈ $150k – $332k per year (Estimated) • New York • Remote (United States, UTC-5 – UTC+1 hours) • PhD • Full-Time
Python
AI/ML
RLHF
Reinforcement Learning
Computer Vision
NumPy
PyTorch
DPO
PPO
GRPO
Post-training
Reward Modeling
Machine Learning
+4
Apply

Cerence

cerence.com
cerence.com • HQ: Burlington, United States • Automotive Software • Conversational AI • Speech & Audio AI • 1001-5000 employees • Est. 2019 • NASDAQ: CRNC • Trust B • Hiring 29
HQ: Burlington, United States • Automotive Software • Conversational AI • Speech & Audio AI • 1001-5000 employees • Est. 2019 • NASDAQ: CRNC • Trust B • Hiring 29
Founded in 2019, Cerence is a global technology company specializing in artificial intelligence and conversational software for the automotive industry. Headquartered in Burlington, Massachusetts, the firm develops voice recognition, natural language understanding, and smart assistant solutions for leading vehicle manufacturers. By integrating intuitive voice controls and connected AI systems into modern automobiles, the organization enhances the overall in-cabin user experience for drivers worldwide.
No named decision-maker yet.
No careers mailbox known
Open 103 daysVerified live · 8 hours ago 3 months ago Long shot · 8%

Senior AI scientist

≈ $24k – $54k per year (Estimated) • Senior • India • Remote (India) • Full-Time • Senior
RLHF
Multimodal AI
Transformers
Mixture of Experts
DPO
GRPO
FSDP
Apply

Prime Intellect

primeintellect.com
primeintellect.com • HQ: San Francisco, United States • Foundation Models • AI Development Tools • Reinforcement Learning • AI Compute & Inference • 201-500 employees • Est. 2023 • Likely startup • Trust A • Hiring 43, 19 new
HQ: San Francisco, United States • Foundation Models • AI Development Tools • Reinforcement Learning • AI Compute & Inference • 201-500 employees • Est. 2023 • Likely startup • Trust A • Hiring 43, 19 new
Prime Intellect is an artificial intelligence infrastructure company headquartered in San Francisco, California, and founded in 2023. The company provides a decentralized platform for training, evaluating, and deploying large-scale AI models, featuring tools for reinforcement learning, agent development, and a global compute marketplace. It operates globally by aggregating computing resources from various providers to enable researchers and developers to build open-source models and autonomous agents.
No named decision-maker yet.
No careers mailbox known
Top 25% payVerified live · 3 hours ago Visa sponsorship 3 months ago Fading · 30%

Applied Research - RL & Agents

$150k – $300k per year • Equity • Hybrid (San Francisco, New York, United States) • Relocation • Full-Time
JavaScript
TypeScript
Databases
Databricks
AI/ML
LangGraph
LangChain
DSPy
Model Context Protocol
vLLM
RLHF
Reinforcement Learning
Function Calling
AI Agents
SGLang
Perplexity
Ray
OpenRouter
Together AI
OpenAI
SFT
GRPO
Post-training
LLM Evaluation
Tool Use
Machine Learning
LLM
Synthetic Data
Frontend
Next.js
React.js
DevOps
Datadog
Prometheus
Cloudflare
Grafana
Terraform
Docker
Kubernetes
Management
Zapier
+28
Apply
Top 25% payVerified live · 3 hours ago Visa sponsorship 3 months ago Fading · 30%

Applied Research - Forward-Deployed

$150k – $300k per year • Equity • Hybrid (San Francisco, New York, United States) • Relocation • Full-Time
Python
JavaScript
TypeScript
Databases
Databricks
AI/ML
LangChain
RLHF
Function Calling
AI Agents
LLM
Perplexity
OpenRouter
Together AI
OpenAI
Hugging Face
SFT
GRPO
Post-training
LLM Evaluation
Tool Use
Reward Modeling
LangGraph
DSPy
Model Context Protocol
Synthetic Data
Frontend
Next.js
React.js
DevOps
Datadog
Cloudflare
Management
Slack
Zapier
+22
Apply

Whitecircle

whitecircle.com
whitecircle.com • HQ: San Francisco, United States • AI Safety & Alignment • 201-500 employees • Est. 2025 • Trust A • Hiring 16
HQ: San Francisco, United States • AI Safety & Alignment • 201-500 employees • Est. 2025 • Trust A • Hiring 16
White Circle is an AI safety company developing best-in-class AI stress-testing and AI moderation tools.
No named decision-maker yet.
Careers mailbox is known
Verified live · 30 min ago 3 months ago Long shot · 23%

Multimodal ML Engineer

≈ $58k – $125k per year (Estimated) • Hybrid (Paris, France, London, United Kingdom) • Relocation • Full-Time
Qwen
DeepSpeed
RLHF
Quantization
Multimodal AI
Knowledge Distillation
AI Agents
PyTorch
Whisper
Mixture of Experts
Synthetic Data
LLaVA
DPO
SFT
GRPO
FSDP
Model Distillation
Reward Modeling
OpenAI
Anthropic
Hugging Face
DevOps
Datadog
QA
Sentry
+15
Apply
Verified live · 30 min ago 3 months ago Long shot · 23%

ML Research Engineer

≈ $55k – $118k per year (Estimated) • Hybrid (Paris, France, London, United Kingdom) • Relocation • Full-Time
Python
SQL
AI/ML
Embeddings
RLHF
NLP
Semantic Search
Hugging Face
DPO
SFT
GRPO
Human-in-the-Loop
Semantic Search
Reward Modeling
OpenAI
Anthropic
DevOps
GitHub
Datadog
QA
Sentry
+10
Apply

Jabil

jabil.com
jabil.com • HQ: Saint Petersburg, United States • Contract Manufacturing • Electronics Manufacturing • 5000+ employees • Est. 1966 • NYSE: JBL • Trust B • Hiring 1794, 199 new
HQ: Saint Petersburg, United States • Contract Manufacturing • Electronics Manufacturing • 5000+ employees • Est. 1966 • NYSE: JBL • Trust B • Hiring 1794, 199 new
Jabil is an American manufacturing solutions company founded in 1966 and headquartered in St. Petersburg, Florida, and one of the largest electronics manufacturing services providers in the world. It designs, prototypes, assembles and ships products on behalf of original equipment manufacturers across cloud infrastructure, healthcare, automotive, capital equipment, packaging and consumer electronics, operating roughly a hundred plants in some thirty countries. Listed on the New York Stock Exchange, it increasingly positions itself as an engineering and supply chain partner rather than a contract assembler, taking on design, regulatory work and aftermarket services alongside production.
Greg Hebard · Chief Financial Officer · decision-maker
Michael Dastoor · Chief Executive Officer · decision-maker
James Golden · Co-Founder · founder
William Morean · Co-Founder · founder
Careers mailbox is known
Open 112 daysVerified live · 3 hours ago 4 months ago Long shot · 4%

SLM & VLM Engineer — Post-Training, Tool Calling & Agents

≈ $55k – $148k per year (Estimated) • In office (Singapore) • Full-Time • Chinese
Python
C++
C++
PyTorch C++
AI/ML
CUDA Toolkit
Quantization
Knowledge Distillation
Function Calling
VLM
PyTorch
CUDA
DPO
SFT
PPO
GRPO
Post-training
Tool Use
Model Distillation
Reward Modeling
+10
Apply
Open 154 daysVerified live · 3 hours ago 5 months ago Long shot · 9%

Senior / Staff SLM & VLM Engineer — Post-Training, Tool Calling & Agents

≈ $96k – $274k per year (Estimated) • Staff+ • In office (Singapore) • Full-Time • Staff • Chinese
Python
C++
C++
PyTorch C++
AI/ML
CUDA Toolkit
Quantization
Knowledge Distillation
Function Calling
VLM
PyTorch
CUDA
DPO
SFT
PPO
GRPO
Post-training
Tool Use
Model Distillation
Reward Modeling
+10
Apply

Inflection AI

inflection.ai
inflection.ai • HQ: Palo Alto, United States • Conversational AI • LLM & Generative AI • 51-200 employees • Est. 2022 • Startup • Trust B • Hiring 6
HQ: Palo Alto, United States • Conversational AI • LLM & Generative AI • 51-200 employees • Est. 2022 • Startup • Trust B • Hiring 6
Inflection AI built the personal artificial intelligence assistant Pi together with its own Inflection model family, trained on one of the largest compute clusters of its era. Most of its founding team and researchers joined Microsoft in 2024 in an unusual arrangement that left the company operating independently. The business now focuses on enterprise deployments of empathetic conversational models rather than frontier scale pretraining; its earlier work on conversational tone and memory shaped several consumer assistant products.
Mustafa Suleyman · Executive Officer, Director · decision-maker
Karen Simonyan · Executive Officer, Director · decision-maker
Reid Hoffman · Director
No careers mailbox known
Open 112 daysTop 25% pay Visa sponsor 4 months ago Long shot · 10%

Principal Research Engineer, Model Training & Post-Training

$400k – $550k per year • Equity • Lead • In office (Palo Alto, United States) • PhD • Principal
Fine-tuning
RLHF
Multimodal AI
Knowledge Distillation
Function Calling
AI Agents
LLM
Synthetic Data
DPO
SFT
GRPO
Post-training
Human-in-the-Loop
Tool Use
Model Distillation
RLAIF
Reward Modeling
Machine Learning
+10
Apply
Machines can read this selection · checked 21:58 UTC
GRPO Jobs - Remote & On-site
Frequently asked questions
How many GRPO jobs are available on Alion?
There are currently 109 active listings from 74 hiring companies on Alion, refreshed daily.
What salary do GRPO jobs pay?
The average salary is around $313,042 per year based on the current listings.
Are remote options available for GRPO jobs?
Yes, many of these roles offer remote or hybrid work, and you can filter listings by remote, relocation and visa sponsorship.
How do I apply for GRPO jobs?
Open any listing on Alion and apply directly with the employer through the provided link - no lengthy sign-up required.