Principal Applied AI Scientist (LLMs) (Hybrid) - Fiddler Labs, Inc

Our Purpose

At Fiddler, we understand the implications of AI and the impact that it has on human lives. Our company was born with the mission of building trust into AI. With the rise of the internet, trust in AI has been degraded by a plethora of issues like spam, fraudulent transactions, hate speech, and online abuse. Fiddler enables organizations to get ahead of these issues by building trustworthy, transparent, and explainable AI solutions.

Fiddler partners with AI-first organizations to help build a long-term framework for responsible AI practices, which, in turn, builds trust with their user base. Data Science, MLOps, and business teams use Fiddler AI to monitor, explain, analyze, and improve their AI solutions to identify performance gaps, mitigate bias, and drive better outcomes. Our platform enables engineering teams and business stakeholders alike to understand the “why” and how behind model outcomes.

Our Founders

Fiddler AI is founded by Krishna Gade (engineering leadership at Facebook, Pinterest, Twitter, and Microsoft) and Amit Paka (two-time founder with acquisitions by Samsung and PayPal and product roles at Expedia and Microsoft). We are backed by Insight Partners, Lightspeed Venture Partners, and Lux Capital.

Why Join Us

Our team is motivated to unlock the AI opaque box and help society harness the power of AI. Joining us means you get to make an impact by helping reduce algorithmic bias and ensure that models in production across many different industries are transparent and ethical. We are an early-stage startup and have a rapidly growing team of intelligent and empathetic doers, thinkers, creators, builders, and everyone in between. The AI and ML industry has a rapid pace of innovation and the learning opportunities here are monumental. This is your chance to be a trailblazer.

Fiddler is recognized as a pioneer in the field of AI Observability and has received numerous accolades, including: 2022 a16z Data50 list, 2021 CB Insights AI 100 most promising startups, 2020 WEF Technology Pioneer, 2020 Forbes AI 50 most promising startups of 2020, and a 2019 Gartner Cool Vendor in Enterprise AI Governance and Ethical Response. By joining our brilliant (at least we think so) team, you will help pave the way in the AI Observability space.

The Opportunity

In this role, you will be able to make a real impact on the safety of large language models and generative AI solutions across different verticals. You will work on the cutting edge of envisioning and building new types of tools and algorithms to monitor, explain, and improve such models.

What You'll Do

• As a Principal Applied AI Scientist, you will play a key role in shaping the future of Fiddler’s observability solutions for Large Language Models (LLMs) and Generative AI. You will develop and tune novel LLMs, ML models, and other techniques to monitor, classify, and protect large language model prompts/responses and related use cases.

• Establish thought-leadership by following up on state of the art research in Large Language models, Gen AI and responsible AI. Conduct experiments, build POCs, evaluate and benchmark novel techniques related to LLM monitoring, debugging, guardrails etc. Identify ways to leverage some of these into product offerings and guide efforts to pursue these.

• Coordinate with multinational teams to drive projects from research POC to production.

• Develop deep product understanding. Think critically and creatively to recognize common customer challenges and craft powerful generalizable solutions.

• Support Sales and Customer Success teams to address some of the more advanced customer questions and asks that require deep understanding of novel ML use-cases. Provide expert guidance to customers with challenging explainability uses.

• Author white papers, blog posts, and develop technical content related to ML and the Fiddler platform.

What We are Looking For

• MS/PhD in Computer Science, Data Science or equivalent discipline.

• 8-10 years of industry experience.

• Deep technical understanding of Large language models. Experience in building, training and fine tuning LLMs. Practical experience with the latest technologies in LLM and generative AI, such as parameter-efficient fine-tuning, instruction fine-tuning, and advanced prompt engineering techniques like Tree-of-Thought / Chain-of-Thought.

• Experience setting up, monitoring and tuning ML models and Large language models in production. Experience with common LLM frameworks (Langchain, Llamaindex, RAG, HuggingFace, BERT models, quantization, and eval frameworks) and MLOps best practices.

• You invest in staying up to date with the LLM space, you regularly educate yourself on new capabilities, market trends and needs and are involved in local user groups and/or events relating to LLMs. You have experience speaking at relevant industry events.

• Proficiency with Python, ability to write production grade code and support ML model deployments that can handle live workloads.

• Ability to conduct deep analysis and benchmarking of models and methods.

• Experience debugging and troubleshooting production systems.

Nice to Have

• Experience working with Computer vision models.

• Experience with data visualization and dimensionality reduction techniques.

• Knowledge of infrastructure management (AWS / Google Cloud, Kubernetes, or equivalent on-premises technologies).

Pay Range

Pay Range in San Francisco California is $250,000 - $320,000. Hybrid with at least 2 days a week in the Palo Alto office, usually Mondays and Thursdays.

The posted range represents the expected salary range for this job requisition and does not include any other potential components of the compensation package, benefits, and perks previously outlined. Ultimately, in determining pay, we'll consider your experience, leveling, location, and other job-related factors.

Equal Opportunity Employer

Fiddler is proud to be an equal-opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees.