Machine Learning Engineer, Ideogram
Applied AI
+ Equity
Generative media tools
Open for applications
Generative media tools
21-100 employees
Open for applications
+ Equity
21-100 employees
Company mission
To make professional-quality design accessible to everyone.
Role
Who you are
- 3+ years of building and shipping ML products
- You write backend code in Python/ PyTorch/ Jax
- You design benchmarks and evaluation metrics, run error analyses, and debug numerical stability across data, training, and inference
- You’ve taken ML features from idea to evaluation to production with measurable gains (quality, latency, or cost)
- You collaborate with the product team to define success criteria and iterate quickly and pragmatically
- You’re comfortable curating, labeling, and cleaning datasets, and designing evaluation corpora and failure-mode analyses
- You’re familiar with modern deep learning such as transformers, diffusion, and encoders
- You bring ownership, urgency, curiosity, and learn fast in a startup setting
What the job involves
- We’re hiring our first Machine Learning Engineer to lead applied AI initiatives. You’ll bridge research and product, turning frontier generative models into production features across our first-party app and API
- You’ll experiment quickly, build rigorous evaluations and datasets, and partner with research, engineering, infra, and product to ship reliable and scalable ML systems
- You’ll fine-tune and deploy models for creative use cases (text-to-image, image-to-text, image enhancement and editing, multimodal), set clear success metrics (quality, latency, cost), and contribute to safety, monitoring, and reliability of our systems
- You’ll lead 0 to 1 projects that shape Applied ML practices at Ideogram, while delivering features that create delight and real value for our users
Our take
With text-to-image software well and truly in the public conscience, Ideogram has launched a platform to support creativity, while capitalising on the popularity of AI image creation services.
Its generative AI platform requires a prompt and rendering style (think anime, fantasy, or cinematic) from which it creates a high quality image for the user. As with other platforms of its type, the images tend to be slightly flawed but interesting, and it appears that almost anything can be created.
Of course, there is a downside to launching into a market that is under such a strong spotlight, especially as many artists raise concerns over their work being used to train image creation AI without their consent. For its part, Ideogram claims to operate with a "high standard for trust and safety", utilizing a $22.3M CAD funding round to grow its team and footprint.
Freddie
Company Specialist at Welcome to the Jungle
Company
Funding (2 rounds)
Feb 2024
$80m
SERIES A
Aug 2023
$16.5m
SEED
Total funding: $96.5m
Company HQ
Financial District, Toronto, ON
Leadership
Mohammad Norouzi
(Co-Founder & CEO)Previously Senior Staff Research Scientist at Google for 7 years.
William Chan
(Co-Founder & CTO)Spent 7 years at Google, ultimately as a Google Brain Staff Research Scientist.
Chitwan Saharia
(Co-Founder)Spent 3 years at Google, ultimately as a Senior Research Scientist. Prior to this, was a Research Assistant at the Montreal Institute for Learning Algorithms.
Jonathan Ho
(Co-Founder)Previously a Research Scientist for Google and for OpenAI.
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 12 more jobs at Ideogram