Senior Machine Learning Engineer, ZoomInfo
MLOps & LLMOps
$180-247.5k
More information about location
Market intelligence and contact details database
Be an early applicant
Market intelligence and contact details database
1001+ employees
Be an early applicant
$180-247.5k
More information about location
1001+ employees
Company mission
To unlock the growth potential of professionals and businesses with a modern end-to-end revenue operating system.
Role
Who you are
- 7+ years building production ML systems (or 4+ post-PhD/MS) with at least two at 100M+ user or record scale
- Deep expertise in modern MLOps/LLMOps (K8s, Docker, MLFlow/Vertex AI/SageMaker, Kubeflow, Airflow, ArgoCD, model registries)
- Proven record optimizing LLMs/SLMs via quantization, LoRA/PEFT, distillation, and distributed training (DeepSpeed, FSDP, Hugging Face, TensorRT-LLM)
- Strong software-engineering fundamentals in Python and experience with distributed computing infrastructure such as Spark and Ray based jobs and pipelines
- Hands-on mastery of vector search & RAG patterns, entity-resolution algorithms, and graph data stores (Neo4j, Neptune, JanusGraph, Spanner Graph) is a major plus
- Fluency in cost, reliability, security, and compliance trade-offs; comfortable providing inputs to budgets and setting SLOs
- Exceptional communication skills—able to align executives, PMs, Data Scientists and engineers around data-driven decisions
What the job involves
- We’re building an elite Applied AI platform that scales from first experiment to planet-scale production—securely, cost-effectively, and fast. Join us as a Senior Machine Learning Engineer and own the MLOps & LLMOps backbone that powers our retrieval, recommendation, and knowledge-graph products
- Architect the end-to-end MLOps / LLMOps stack (Kubernetes, Ray, Argo, Terraform, MLFlow, Feature & Model Stores) for training, fine-tuning, and serving LLMs, RAG pipelines, NER, and entity-resolution models at multi-billion-record scale
- Design cost-aware training & inference workflows—automatic mixed precision, quantization, distillation, dynamic batching, on-demand GPU/CPU autoscaling—to cut $/call and CO₂ while meeting 99.9% SLAs
- Build iron-clad evaluation & safety frameworks: red-teaming, hallucination scoring, PII/toxicity filters, guardrail policies, shadow deployments, and continuous regression tests
- Fine-tune embedding models and pre-train domain-specific LLMs on ZoomInfo corpora; integrate with vector databases (Pinecone, Milvus, OpenSearch) to power RAG search and recommender systems
- Prototype and benchmark emerging AI/infra tech (parameter-efficient tuning, retrieval frameworks, serverless GPUs) against incumbent solutions; present clear “buy-build-borrow” recommendations
- Model, build, and optimize knowledge graphs—schema design, entity linking, incremental updates—to supercharge retrieval and recommendation accuracy
- Profile and re-architect RAG, NER, and entity-resolution services for sub-100 ms latency, 5× QPS jumps, and zero-downtime deploys across global regions
- Lead multi-team CI/CD best practices—blue/green, canary, GitOps—and large-scale A/B tests to quantify revenue impact
- Mentor engineers, publish internal playbooks & external blogs, and speak at industry forums to cement ZoomInfo’s technical leadership
Our take
It was only a few years ago when B2B sales teams relied on warm introductions and contact books. Now, cloud-based CRM, automation and Big Data are critical tools for any selling process. ZoomInfo’s B2B contact base has played a massive part in that journey.
The company primarily helps salespeople find potential customers. Crucially, ZoomInfo then leverages technology to help make that approach as successful as possible. Think contact details of the most likely budget holder at that company, signal for when and how to approach and crucial updates on the company’s fortunes.
ZoomInfo has grown significantly over the previous few years and is now shifting its approach to pursue more inorganic growth through mergers and acquisitions, acquiring several companies in 2022, including the job recruitment website Comparably.
The risk for ZoomInfo is that these moves represent expansion beyond its traditional strengths in top-of-funnel customer acquisition. The company will need to ensure that it maximizes synergies in order to win in these markets. The company has expanded operations in India and is set to see further expansion, as indicated by its appearance on the Fortune Future 50 list of robust and growing companies in 2023.
Freddie
Company Specialist at Welcome to the Jungle
Insights
7% employee growth in 12 months
Company
Funding (1 round)
Jul 2004
$7m
SERIES A
Total funding: $7m
Company benefits
- Benefits for the Best You - We want our employees and their families to thrive. In addition to comprehensive benefits we offer holistic mind, body and lifestyle programs designed for overall well-being
- Base Pay - We attract and retain the best talent with competitive compensation that aligns with industry benchmarks and merit programs yearly for promotions and salary increases based on performance, time in role, and position to market
- Bonus - When you crush it, we like to celebrate! Bonuses are awarded annually in connection with company and individual performance
- Equity - We offer each employee a generous equity award in the form of RSUs, so they can directly feel the results and success of their hard work
- Vacation & Time Off - We know everyone's needs are different so we offer flexible paid time off plans including an unlimited vacation option for many employees
- Hybrid Working Model - We offer flexible hybrid remote working models to promote equity, employee wellness, and work life balance
Company HQ
Vancouver
Leadership
Henry Schuck
(CEO)Originally founded DiscoverOrg which went on to become ZoomInfo through an acquisition in 2019.
Salary benchmarks
We don't have enough data yet to provide salary benchmarks for this role.
Submit your salary to help other candidates with crowdsourced salary estimates.
Share this job
View 50 more jobs at ZoomInfo