MLOps Engineer — JD

Multimodal is a leading generative AI enterprise software company based in New York City. Our mission is to automate complex knowledge tasks and workflows for enterprises in the banking, insurance, and healthcare sectors. We develop custom AI agents that use state-of-the-art large language models to automate manual, repetitive knowledge-based workflows involving documents and databases.

Our common applications include document classification and processing (" Document AI "), decision making using multiple data streams (" Decision AI "), retrieving information from documents and databases for end users via a conversational interface (" Conversational AI " and " Database AI "), and generating new content based on company-specific data (" Content AI ").

By integrating AI into our clients' workflows, we help them modernize their workforce and achieve enterprise-wide automation with superhuman accuracy and near-real-time decision making .

We pride ourselves in being fast (delivering ROI in as little as 3 months), effective (surpassing human-level performance), and reliable (partnering with our clients for the long term).

We are a team of product, engineering, sales, and marketing professionals, many of whom have founded and operated multiple ML startups over the years. We invite you to join us on this journey to bring the latest in AI and NLP to more companies and products.

As a MLOps engineer at Multimodal, you will design and develop solutions for clients across many use cases and many verticals, leveraging state-of-the-art large language models.

Explore client’s product vision and needs

Evaluate possible approaches to solving the problem and decide on best course of action (best defined as fast, economical, and effective)

Work with our internal product team to develop both consumer-facing and business-facing products

Design, implement and maintain automated CI/CD pipelines for ML models

Design and implement model deployment and monitoring strategies

Work closely with our machine learning and engineering teams to improve the efficiency and scalability of our ML workflows

Develop techniques to train and serve models faster

Deploy models as API

Proficiency in Python and related ML frameworks such as PyTorch and Tensorflow

Proficiency with large language models such as OpenAI’s GPT-3.5, GPT-4, Meta’s Llama 1 and Llama 2, Vicuna, Koala, Wizard, Falcon, RedPajama, Anthropic’s Claude 1 and Claude 2, GPT-J, GPT-NeoX, T5, Flan-UL2, and others

Familiarity with autoregressive sequence models, such as Transformers

Experience with Hugging Face, Weights & Biases, LangChain, Pinecone, and similar libraries

Experience using large-scale distributed training strategies

Experience with writing production-ready code, using CI/CD, and deploying models on major cloud platforms such as AWS, GCP, and Azure and specialty GPU players such as Coreweave, Lambda Labs, Modal Labs, and OctoML

Experience using Kubernetes and Docker

Awareness of and experience with specialized hardware providers such as Cirrascale, Cerebras, SambaNova, and GraphCore

Experience using GPUs such as A100s and H100s

Strong written and oral communication and problem-solving skills

A demonstrated passion for applied NLP models and products

Passion for learning and applying the latest in LLM and NLP research