Careers
Build the open AI stack with us.
We're a small team of researchers and engineers building inference, storage, and fine-tuning that anyone can use — and publishing the models and tools we make along the way. If that mission resonates, we'd love to work with you.
Open roles are listed below. We're remote-friendly and hire for range as much as depth — if you're excellent and excited about open AI infrastructure, reach out even if nothing is an exact match.
Research
4 openResearch Scientist, Pre-Training
You'll help design and train Permissionless Labs' open-weight foundation models from first principles — making the architecture, data, and scaling decisions that define what we release to the world.
What you'll do
- Run pre-training experiments on architecture, data mixtures, and optimization
- Develop and validate scaling laws to guide larger training runs
- Work with infrastructure to improve training efficiency and stability
- Help decide what we open-source and document it for the community
What we're looking for
- Experience training transformer models at scale
- Strong grounding in deep learning and large-scale optimization
- A bias toward rigorous, reproducible experimentation
- Enthusiasm for publishing open weights and open science
Research Scientist, Post-Training
You'll turn capable base models into genuinely useful ones — owning fine-tuning, RLHF, and preference optimization for the open models we ship.
What you'll do
- Design and run supervised fine-tuning and RLHF pipelines
- Build reward models and preference datasets
- Improve instruction-following, safety, and reasoning behavior
- Measure quality with rigorous evals and ship the improvements
What we're looking for
- Hands-on experience with post-training methods (SFT, RLHF, DPO, etc.)
- Comfort working across data, training, and evaluation
- Good judgment about model behavior and where it breaks
- Care for openness and reproducibility
Research Scientist, Inference & Efficiency
You'll make our models fast and cheap to run without giving up quality — the difference between a great model and a great product.
What you'll do
- Research quantization, distillation, and speculative decoding
- Co-design the model and serving stack for throughput and latency
- Benchmark quality-versus-speed trade-offs honestly
- Ship efficiency wins into the PipeNetwork.AI serving path
What we're looking for
- Deep understanding of transformer inference
- Experience with quantization or model compression
- Systems intuition for where latency and cost really come from
- A pragmatic, measurement-driven approach
Research Engineer, Evaluation
You'll build the evals and benchmarks that decide whether a model is good enough to release — and keep them honest as the models get better.
What you'll do
- Design task-specific and general evaluation suites
- Build reproducible eval harnesses and dashboards
- Investigate regressions and surprising results
- Push back on metrics that don't reflect real quality
What we're looking for
- Strong software engineering plus ML literacy
- Healthy skepticism and attention to measurement detail
- Experience evaluating LLMs or ML systems
- Clear technical writing
ML Infrastructure
4 openResearch Engineer, Inference Systems
You'll build the serving engine behind PipeNetwork.AI — the system that turns model weights into a fast, reliable API.
What you'll do
- Build and optimize the high-throughput inference server
- Implement batching, KV-cache management, and scheduling
- Drive down time-to-first-token and cost-per-token
- Work with research to land new model architectures
What we're looking for
- Strong systems programming (C++, Rust, CUDA, or Python)
- Experience with GPU inference or high-performance serving
- Care about latency, throughput, and reliability
- Ability to reason from first principles about performance
Research Engineer, GPU Kernels
You'll write the low-level kernels that make our models fast on real hardware — where a few percent matters at fleet scale.
What you'll do
- Write and tune CUDA and Triton kernels for training and inference
- Profile and eliminate bottlenecks across the stack
- Implement fused and low-precision operations
- Contribute kernels back to open source where we can
What we're looking for
- Experience writing CUDA or Triton kernels
- Deep understanding of GPU architecture and memory hierarchies
- An obsession with profiling and measured wins
- Systems-level rigor
Research Engineer, Training Systems
You'll own the distributed training stack our largest runs depend on — parallelism, throughput, and rock-solid reliability.
What you'll do
- Build and scale distributed training (data, tensor, and pipeline parallel)
- Make long training runs fault-tolerant and reproducible
- Optimize throughput and hardware utilization
- Partner with research on new training workloads
What we're looking for
- Experience with large-scale distributed training
- Familiarity with PyTorch internals and collective communication
- Comfort debugging across the hardware/software stack
- A reliability mindset
Research Engineer, Numerics
You'll guard correctness and reproducibility across the stack — especially where low precision and performance work threaten to break things quietly.
What you'll do
- Investigate and fix numerical correctness issues
- Validate low-precision training and inference paths
- Build tooling to catch silent regressions
- Set standards for reproducibility across the team
What we're looking for
- Strong grasp of floating-point and numerical methods
- Experience with mixed- or low-precision ML
- Patience for subtle, high-stakes bugs
- Careful, methodical engineering
Engineering & Platform
6 openSoftware Engineer, Inference Platform
You'll build the product around our inference engine — the APIs, autoscaling, and reliability that make PipeNetwork.AI something people trust in production.
What you'll do
- Build and operate the inference API and control plane
- Implement autoscaling across GPU fleets
- Add observability, quotas, and reliability safeguards
- Keep the developer experience clean and OpenAI-compatible
What we're looking for
- Strong backend and distributed-systems experience
- Comfort operating production services
- Product sense for developer APIs
- Ownership and pragmatism
Software Engineer, Storage & Data Infrastructure
You'll build the content-addressed, versioned storage that models and datasets live in — durable, fast, and with a full provenance trail.
What you'll do
- Build content-addressed, deduplicated object storage
- Implement versioning, lineage, and fast retrieval
- Integrate storage tightly with training and inference
- Scale to petabytes without losing durability
What we're looking for
- Experience with distributed storage or data systems
- Strong systems and backend engineering
- Care about durability, consistency, and performance
- Comfort operating at large data scale
Software Engineer, Fine-Tuning Platform
You'll build the managed fine-tuning product that takes a customer from a dataset to a deployed model in a few steps.
What you'll do
- Build the fine-tuning pipeline and job orchestration
- Integrate evaluation into the training loop
- Ship one-step deploy from tuned weights to inference
- Make the whole flow reliable and easy to use
What we're looking for
- Strong backend engineering and orchestration experience
- Familiarity with ML training workflows
- A product focus on developer experience
- Reliability mindset
Software Engineer, Full Stack
You'll build the surfaces developers actually touch — the dashboard, docs, and console across our inference, storage, and fine-tuning products.
What you'll do
- Build front-end and back-end for the developer console
- Design clean, fast UI for complex workflows
- Improve docs, onboarding, and self-serve flows
- Partner with platform teams end to end
What we're looking for
- Strong full-stack experience (TypeScript/React plus a backend)
- An eye for clean, usable interfaces
- Ability to own features end to end
- A bias to ship
Site Reliability Engineer (SRE)
You'll keep GPU fleets and serving infrastructure healthy, observable, and calm — so builders can rely on us.
What you'll do
- Own reliability, monitoring, and incident response
- Build automation for GPU fleet operations
- Improve observability across the platform
- Drive down toil and make on-call humane
What we're looking for
- Strong SRE or production-operations experience
- Comfort with cloud, containers, and infrastructure-as-code
- Calm, systematic incident response
- An automation-first mindset
Software Engineer, Security
You'll harden the platform and protect what matters most to our customers — their data, their weights, and their workloads.
What you'll do
- Build security into the platform and infrastructure
- Protect customer data, models, and isolation boundaries
- Lead threat modeling and security reviews
- Respond to and prevent security issues
What we're looking for
- Strong security-engineering background
- Experience securing cloud infra and multi-tenant systems
- Practical, risk-based judgment
- A collaborative, builder mentality
Product & Developer Relations
2 openProduct Manager, Developer Platform
You'll own the roadmap for our developer products — deciding what we build across inference, storage, and fine-tuning, and why.
What you'll do
- Own product strategy and roadmap for the platform
- Talk to developers and turn their needs into a plan
- Partner with engineering to ship and iterate
- Define success and measure it honestly
What we're looking for
- Experience as a PM for technical or developer products
- Strong technical fluency
- Clear prioritization and communication
- Customer obsession
Developer Advocate
You'll help developers succeed with our platform — through docs, demos, and open-source examples that actually work.
What you'll do
- Write docs, tutorials, and sample applications
- Build and maintain open-source examples and SDK demos
- Gather developer feedback and advocate for it internally
- Represent us in the open-source and AI communities
What we're looking for
- Strong engineering plus great technical writing
- Experience with developer tools or DevRel
- Genuine enthusiasm for open source
- Empathy for the developer experience
Operations
2 openTechnical Recruiter
You'll help us find and hire the researchers and engineers who will build the open AI stack — one of the highest-leverage things we do.
What you'll do
- Own technical sourcing and full-cycle recruiting
- Build a great, humane candidate experience
- Partner with hiring managers on what 'great' looks like
- Help shape how we scale the team
What we're looking for
- Technical recruiting experience, ideally in ML or infrastructure
- Strong sourcing and candidate judgment
- Excellent communication
- Care about people and mission
Chief of Staff
You'll keep a fast-moving lab aligned, resourced, and focused on what matters — working closely with leadership across the whole company.
What you'll do
- Drive planning, priorities, and follow-through
- Run key operating rhythms and communications
- Take on high-leverage projects end to end
- Connect research, engineering, and operations
What we're looking for
- Experience in operations, strategy, chief-of-staff, or founding roles
- Exceptional organization and judgment
- Comfort with ambiguity and speed
- Low ego and high ownership
Don't see the right role? We're always glad to meet people who care about open, accessible AI infrastructure.
Send a general application