Founding Developer Relations Engineer

Goaly is emerging from stealth with a focus on post-training efficiency. As the founding Developer Relations Engineer, you will translate real technical work into benchmarks, reproducible tooling, and practitioner-ready documentation, while building distribution across the ML infrastructure community.

Role Overview

This is Goaly’s first Developer Relations hire and the technical voice for the company as it comes out of stealth. Users include ML infra engineers and researchers who run post-training workloads on GPU clusters. They rely on benchmarks, repositories, and technical documentation more than marketing.

The role is split roughly evenly: half of your time building and validating real post-training workloads on Goaly’s stack, and the other half ensuring practitioners find and adopt the artifacts through content, talks, open-source releases, and direct practitioner conversations that can lead to design partners and hires.

Key Responsibilities

• Build the signature technical artifacts, including a public post-training efficiency benchmark, reproducible before-and-after case studies (including doomed-run kills, silent throughput regressions, and dollars per eval point), plus the open-source tools released around these artifacts.
• Write conversion-focused documentation such as quickstarts, reference docs, and example integrations for practitioner stacks including PyTorch distributed, Megatron, FSDP, TorchTitan, Ray, Slurm, Kubernetes on GPUs, Weights & Biases (W&B), and MLflow. The goal is enabling engineers to generate their own read-only Retroactive Kill Report on their cluster without contacting Goaly.
• Create technical content practitioners share: deep-dive engineering posts, benchmark write-ups, failure-mode teardowns (for example, “seven ways an RL run dies silently”), short demo videos, and live walkthroughs.
• Own distribution by leading Goaly’s technical presence on X, LinkedIn, GitHub, Hacker News, and ML-infra communities such as the ML-infra Discord and Slack. Lead outreach through talks at venues practitioners attend including PyTorch Conference, MLSys, Ray Summit, GTC, and NeurIPS and ICML workshops, plus Bay Area AI infrastructure meetups.
• Operate as the feedback loop by pairing with design partners, reproducing their problems on Goaly’s stack, and delivering precise prioritized feedback to engineering. Convert repeated questions into documentation, tools, or product changes.
• Build community around the problem space through practitioner conversations, office hours, and a newsletter or series focused on post-training efficiency.
• Amplify the founders by turning the founding team’s Llama post-training infrastructure experience into content and talks, and partnering with the GTM team to support a design-partner pipeline and candidate pipeline.
• Measure and improve distribution and adoption by tracking reach, repository adoption, documentation funnel metrics, inbound design-partner requests, and hires sourced. Double down on what works and reduce investment in what does not.

Required Qualifications

• 3+ years as a software or ML engineer, with evidence you still write code weekly.
• An engineering background that includes the ability to stand up a distributed training job, interpret profiler traces, and debug CUDA OOM issues without help.
• Hands-on experience with LLM post-training and/or the infrastructure beneath it, including SFT, RL, LoRA, DPO, or systems such as PyTorch distributed, Megatron, DeepSpeed, FSDP, Ray, Slurm, and GPU Kubernetes.
• A track record of explaining complex topics clearly through blog posts, documentation, conference talks, open-source READMEs, tutorials, or videos (sharing links as part of the application process).
• Demonstrable distribution capacity, such as an audience on X, LinkedIn, YouTube, a newsletter, or GitHub, plus talks and community presence, or a history of building reach from zero (with reach in ML infrastructure and post-training communities valued over follower count).
• Two-way communication skills, including trusted practitioner-facing writing and the ability to listen to infrastructure engineers and return with clear engineering priorities.
• A preference for substance: favoring one reproducible benchmark over multiple hot takes.
• Comfort being the public face of an early-stage effort where there is no fixed playbook, messaging may shift, and deliverables have your name attached.

Technical Skills and Tools

• PyTorch distributed, Megatron, FSDP, TorchTitan, Ray, Slurm, Kubernetes on GPUs
• Weights & Biases (W&B), MLflow
• SFT, RL, LoRA, DPO, DeepSpeed
• CUDA OOM
• W&B, Nsight, PyTorch Profiler

Benefits

• Hybrid role in Palo Alto: 4+ days per week in office, with roughly 20% travel for conferences and customer visits.
• Visa sponsorship: H-1B and OPT/CPT support available, with immigration counsel.
• Complimentary lunch, dinner, snacks, and drinks.

How You Will Work

• Mission first: choose work based on impact and take responsibility for outcomes.
• High agency: identify gaps, form a plan, and move without waiting for perfect clarity.
• Speed with rigor: ship, measure, and iterate quickly while protecting correctness, safety, and reliability.
• Flexible scope: collaborate across teams and technical boundaries when it is the fastest path to the real problem.
• Low ego, high standards: provide direct feedback, update decisions with new evidence, and help the team win.
• Continuous learning: the stack changes quickly, so expect to learn unfamiliar systems and domains as the work requires.