Full-Time

Staff Machine Learning Engineer

Instruction-tuning & Alignment

Updated on 12/6/2024

Scribe

Scribe

51-200 employees

AI-powered platform for creating training documents

Consumer Software
Enterprise Software
AI & Machine Learning

Compensation Overview

$190k - $250kAnnually

+ Equity + Benefits

Senior

San Francisco, CA, USA

Hybrid position.

Category
Applied Machine Learning
AI & Machine Learning
Required Skills
Python
Jupyter
Requirements
  • You have a relevant degree from a top ML program — minimum 5+ years of industry experience working deep in the weeds on hard ML problems.
  • You have strong software engineering skills (including Python, Jupyter, etc.)
  • You have experience with instruction-based finetuning and RLHF.
  • You have experience deploying LLMs in production.
  • You have great intuitions of fundamental ML concepts (e.g., fluent in thinking about overfitting, generalization, back-propagation, etc.)
  • You have strong communication skills.
  • You have a track record of successfully owning projects from start to finish.
Responsibilities
  • Explore cutting-edge techniques in the artificial intelligence field and translate these innovations into valuable features for our millions of users.
  • Participate to the creation of the most qualitative enterprise software workflow dataset.
  • Perform multimodal product research and optimize our data flywheel to improve the performance of our models.
  • Fine-tune and train large language models leveraging various modalities.
  • Be responsible for the alignment of our state-of-the-art proprietary models so that they can behave as intended for our users.
  • Collaborate with product teams to ensure your work translates into better experiences for Scribe users (instruction following, guard-railing, etc.)

ScribeHow offers a platform that uses artificial intelligence to create how-to guides and training documents automatically. This service helps businesses save time by generating step-by-step instructions for any web or desktop process and redacting sensitive information from screenshots for compliance. Operating on a freemium model, ScribeHow attracts a wide range of customers, from small businesses to large corporations. The goal is to simplify onboarding and training processes, enabling businesses to scale effectively.

Company Stage

Series B

Total Funding

$53.5M

Headquarters

San Francisco, California

Founded

2019

Growth & Insights
Headcount

6 month growth

17%

1 year growth

43%

2 year growth

204%
Simplify Jobs

Simplify's Take

What believers are saying

  • ScribeHow's recent $25M Series B funding round indicates strong investor confidence and provides resources for further innovation and growth.
  • The platform's ability to help businesses scale rapidly, as evidenced by customer testimonials, highlights its effectiveness and potential for widespread adoption.
  • Continuous feature updates, such as the recent launch of Scribe AI, demonstrate the company's commitment to innovation and improving user experience.

What critics are saying

  • The competitive SaaS market for documentation tools requires ScribeHow to continuously innovate to maintain its edge.
  • Reliance on AI for documentation may face challenges if the technology fails to accurately capture complex processes.

What makes Scribe unique

  • ScribeHow leverages AI to automatically generate step-by-step guides, eliminating the need for manual documentation, unlike traditional documentation tools.
  • The platform's automatic redaction of sensitive information from screenshots ensures compliance with data protection regulations, a feature not commonly found in competitors.
  • ScribeHow's freemium model attracts a broad range of customers, from small businesses to large corporations, providing flexibility and scalability.

Help us improve and share your feedback! Did you find this helpful?