Founding Machine Learning Engineer
Founding ML engineer (2nd ML hire) at Adaptive, a cybersecurity startup using LLMs and ML models to detect and respond to AI-powered cyberattacks in real time. You build the ML function from scratch — strategy, infrastructure, data pipelines, training, evaluation, and inference serving — coding in Python/Java/TypeScript with PyTorch, Spark, and cloud ML platforms.
- Compensation
- Not disclosed
- Location
- NYC
- Work setup
- Onsite
- Seniority
- Not specified
- Employment
- Not specified
- Experience
- 4+ years
- Category
- Ml / AI
- Posted
Skills
Must have
- 4+ years of experience building ML systems in production
- Strong software engineering fundamentals; write production-quality code in modern languages (Python, Java, TypeScript) and work within large codebases
- Experience with cloud ML infrastructure (AWS SageMaker, Bedrock, Modal, Baseten, or similar)
- Experience with common ML and data processing frameworks (PyTorch, TensorFlow, Spark)
- Comfortable working across the stack — infrastructure, backend services, and data systems
- High autonomy — expected to define the path forward and drive it
Nice to have
- Ideally experience standing up the ML function at an early-stage startup or as the senior/lead ML person at a previous company
Full description
About Adaptive
NVIDIA and OpenAI’s only AI cybersecurity investment.
Adaptive is a cybersecurity startup on a mission to stop AI-powered cyberattacks. In December 2025, the company announced an $81M Series B led by NVIDIA and Bain Capital Ventures, with participation from Capital One Ventures, Citi Ventures, and continued support from Andreessen Horowitz (a16z), the OpenAI Startup Fund, and Abstract Ventures. The round marked NVIDIA’s first AI cybersecurity investment.
Adaptive was founded by Brian Long and Andrew Jones, repeat entrepreneurs who have built and scaled category-defining companies. Brian and Andrew previously co-founded Attentive, which grew to more than $500M in annual revenue and a $10B+ valuation, and TapCommerce, which was acquired by Twitter. Together, they bring deep experience building high-growth, product-led businesses at massive scale as Adaptive builds the security layer for the AI era.
Trusted by leading banks, technology companies, and healthcare organizations, Adaptive protects teams from emerging threats like deepfakes, smishing, and AI-powered voice scams. With rapid enterprise adoption and a $200B+ market ahead, the company is just getting started.
Role
We are seeking our 2nd ML Engineer to define and build Adaptive's ML capabilities. Our products use LLMs and ML models to detect, classify, and respond to cybersecurity threats in real time. The ML problems here are adversarial by nature. Attackers actively evolve to evade detection, labeled data is scarce for novel attack vectors, and models need to operate at production scale with strict latency constraints.
We don't have dedicated ML infrastructure or an ML team today. You'll be building this from the ground up. You will set the technical direction, stand up the infrastructure, and do the hands-on work yourself. You’ll be instrumental in growing the function as our ML surface area expands this year and beyond.
Compensation
Adaptive has raised $146M from top-tier investors including BCV, a16z, NVIDIA, and OpenAI. Our roadmap depends on standing up an ML function that doesn’t yet exist, and we’re ready to compensate accordingly. Expect market level compensation and huge ownership over the ML platform from day one.
Responsibilities
- Define Adaptive's ML strategy: where ML should be applied across our products, what infrastructure we need, and how we should approach build vs. buy decisions.
- Design and build production ML systems end-to-end — data pipelines, model training, evaluation frameworks, and inference serving.
- Establish evaluation methodology. Define how we measure model quality, catch regressions, and make data-driven decisions about model changes.
- Own the strategy for getting the data you need, in the format you need it — what/how to label, how to build feedback loops, and how our models improve over time.