LavX News · May 2026
Secondary coverage of Cedana’s hiring and deployment motion, highlighting kernel-level GPU checkpointing for research labs, enterprise clusters, and hybrid environments.
Fast, reliable, reproducible AI with GPU live migration
Company research
Public company, workplace, funding, and market signals
Updated Jul 30, 2026
Overview
Cedana is a New York-based, fully remote AI infrastructure company founded in 2023 that provides save/migrate/resume infrastructure for CPU and GPU workloads, enabling live checkpointing, migration, and recovery across nodes, clusters, and clouds.
Primary product
GPU/CPU live migration and save/migrate/resume infrastructure
Founded
2023
Headquarters
New York, New York, United States
Team size
10-20
Work style
Remote
Industry
Software Development
Sub-industry
AI infrastructure / workload orchestration
Offices
New York City, New York, United States (headquarters)
United States (distributed workforce)
India (distributed workforce)
Brazil (distributed workforce)
Canada (distributed workforce)
Business model
open-source software
managed service
quote-based enterprise sales
Funding
Stage
pre-seed
Total raised
$500K
Latest round
Pre-Seed Convertible Note · Sep 2023
Latest amount
$500K
Sep 2023 · Y Combinator
People & investors
Work & culture
Fully remote and distributed, with in-person team meetups roughly every 6-8 weeks. The public material emphasizes deep technical work, continuous learning, documentation, teaching, and low-level systems engineering across the full stack.
Work style
Remote
Visa sponsorship
Limited
Compensation
Public job posts show US-remote roles paying $140k-$180k base salary plus meaningful early-stage equity, 0.10%-0.30% equity on some roles, and benefits including fully covered medical, dental, and vision, unlimited PTO, a 401(k), and a workstation budget. Hiring language indicates US citizenship/visa eligibility requirements, so sponsorship appears limited.
Benefits
100% covered medical, dental, and vision insurance for employees and families
Unlimited PTO
401(k) plan
Workstation setup budget
Meaningful early-stage equity
Product & market
Pricing
Not publicly listed; sales-led and quote-based, with both an open-source package and a managed service. The contact page invites users to discuss pricing and request demos.
Differentiators
- —Save/Migrate/Resume for CPU and GPU workloads across instances, nodes, clusters, and vendors
- —Zero code changes and no manual checkpoint hooks required
- —Kernel/OS-level approach with CRIU and low-level runtime integration
- —Live migration for stateful workloads rather than restart-based recovery
- —Claims of up to 80% cost savings and 2-10x faster time-to-first-token
- —Integrations with Kubernetes, Kueue, Slurm, and KServe
- —Works as both an open-source package and a managed service
Technology
Go
Python
Linux
Linux kernel
Kubernetes
Slurm
Kueue
CRIU
Docker
QEMU
KVM
libvirt
Helm
Terraform
NVIDIA CUDA
NVIDIA GPU Driver API
Postgres
RabbitMQ
Kata Containers
Cloud Hypervisor
Customers
Caltech
Fortune 100 pharma lab
AI training teams
AI inference providers
HPC and scientific computing labs
computational biology teams
MLOps platforms
neoclouds / infrastructure providers
enterprise GPU cluster operators
research universities and labs
Competitors
Slurm
Kubernetes-based job orchestration stacks
Lambda Labs
NVIDIA CUDA checkpoint/restore
in-house workload migration tooling
Scale & contact
Estimated revenue
≈$1.4M estimated ARR (third-party estimate, 2025)
Latest news
Cedana Substack · Jun 2025
Explains how Cedana integrates with Kueue for scientific computing, using system-level checkpointing and live migration to preempt workloads safely and resume them without losing progress; includes a Caltech use case.
Cedana Substack · May 2025
Details Cedana’s GPU driver/API interposition approach, benchmark claims, minimal overhead, faster inference throughput, and order-of-magnitude faster restored cold starts.
Northflank · May 2025
A customer case study describing Cedana’s fully remote, engineer-heavy team, production deployments, secure runtime testing, customer delivery model, SOC 2 progress, and emphasis on live migration for GPU-heavy workloads.
Cedana Substack · Nov 2024
A hands-on guide showing Cedana checkpointing and restoring stateful workloads across spot interruptions, with use cases spanning data science, finance, and scientific simulation.
Jobs and careers at Cedana
0 jobs at Cedana
No open positions
Check back later for new openings