- Company Name
- Reducto
- Job Title
- Machine Learning Engineer
- Job Description
-
**Job Title:** Machine Learning Engineer
**Role Summary:**
Design, train, fine‑tune, and deploy vision and multimodal language models that parse complex enterprise documents (PDFs, spreadsheets, images). Build end‑to‑end data pipelines, evaluate model accuracy, and integrate production‑ready components into a scalable product. Work closely with founders and customers to shape model strategy and iterate rapidly based on quantitative results.
**Expectations:**
- Deliver high‑quality models within tight timelines, ensuring no compromise on accuracy.
- Own the full lifecycle: ideation, experimentation, debugging, and shipping.
- Collaborate with product and engineering teams to translate customer needs into measurable model improvements.
- Maintain rigorous version control, documentation, and reproducibility.
- Continuously monitor model performance and propose enhancements.
**Key Responsibilities:**
- Train and fine‑tune state‑of‑the‑art computer vision and vision‑language models on large, unstructured data sets.
- Build and maintain data ingestion, preprocessing, and annotation pipelines for PDFs, tables, and images.
- Design and execute experiments to improve LLM accuracy on extraction, summarization, and form‑completion tasks.
- Integrate models into the production stack, exposing services via APIs or lightweight web interfaces (e.g., Streamlit).
- Evaluate model performance using metrics (F1, precision/recall, mAP, BLEU, etc.) and present insights to stakeholders.
- Iterate fast on feedback loops, troubleshooting issues in production.
**Required Skills:**
- 2+ years of production ML/AI experience.
- Strong Python programming; expertise in PyTorch/TensorFlow or equivalent.
- Deep knowledge of computer vision techniques (OCR, layout analysis, table/graph extraction).
- Experience with vision‑language models (CLIP, PaLI, YoloSeq, etc.).
- Familiarity with data pipeline tools (Pandas, Dask, Airflow) and model serving (FastAPI, Docker, Kubernetes).
- Proficiency in designing experiments, A/B testing, and statistical evaluation.
- Comfortable building lightweight tools (Streamlit, Gradio) for prototyping.
- Self‑motivated, high quality standards, and ability to ship fast without sacrificing correctness.
**Required Education & Certifications:**
- Bachelor’s degree in Computer Science, Electrical Engineering, Applied Mathematics, or related field (Master’s or PhD preferred).
- Relevant certifications (e.g., TensorFlow Developer, AWS Certified Machine Learning) are a plus.
San francisco, United states
On site
Junior
20-11-2025