- Company Name
- Ideogram
- Job Title
- Machine Learning Engineer, Applied AI
- Job Description
-
**Job Title**
Machine Learning Engineer, Applied AI
**Role Summary**
Lead the application of cutting‑edge generative AI models into production features for a creative design platform. Translate research‑grade diffusion, transformer, and multimodal models into scalable, low‑latency, cost‑effective backend services that power text‑to‑image, image‑to‑text, and image enhancement APIs. Own end‑to‑end ML lifecycle—from data curation and training to evaluation, deployment, and monitoring—while collaborating closely with product, engineering, and infra teams to deliver measurable improvements in quality, latency, and operational efficiency.
**Expectations**
- Deliver production‑ready ML systems that show clear gains in quality, latency, or cost.
- Operate with high ownership, rapid iteration, and cross‑functional collaboration.
- Lead 0‑to‑1 AI initiatives and shape applied AI best practices.
- Maintain rigorous safety, monitoring, and reliability standards for deployed models.
**Key Responsibilities**
- Design, develop, and maintain backend ML services in Python using PyTorch or JAX.
- Curate, label, and clean dataset pipelines; create balanced evaluation corpora and failure‑mode analyses.
- Build benchmarks, define success metrics (e.g., FID, precision/recall, latency, cost), and run error analyses.
- Fine‑tune generative models for multimodal creative use cases and deploy them to production APIs.
- Collaborate with product to set success criteria and iterate features through short release cycles.
- Debug numerical stability across training and inference; optimize for throughput and latency.
- Monitor model performance post‑deployment, implement safety checks, and drive continuous improvement.
- Lead cross‑team efforts to integrate infrastructure, scaling, and observability into ML workflows.
**Required Skills**
- 3+ years experience building and shipping ML products.
- Strong programming in Python; deep familiarity with PyTorch or JAX.
- Expertise in modern deep‑learning architectures: transformers, diffusion models, multimodal encoders.
- Proven ability to design evaluation metrics, run benchmarks, and conduct error analyses.
- Experience with data curation, labeling pipelines, and dataset versioning.
- Solid understanding of scaling ML models for production (batching, quantization, GPU/TPU inference).
- Excellent communication and collaboration skills; ability to translate ML concepts to product stakeholders.
- Comfortable with rapid prototyping, A/B testing, and data‑driven decision making.
**Required Education & Certifications**
- Bachelor’s or Master’s degree in Computer Science, Electrical Engineering, Data Science, or related field.
- Relevant certifications (e.g., TensorFlow Professional, PyTorch Advanced) are a plus but not mandatory.