Dow Career Experiences for PhDs – Optical Chemical Structure Recognition
About Dow At Dow, we believe in putting people first and we’re passionate about delivering integrity, respect and safety to our customers, our employees and the planet. Our people are at the heart of our solutions. They reflect the communities we live in and the world where we do business. Their diversity is our strength. We’re a community of relentless problem solvers that offers the daily opportunity to contribute with your perspective, transform industries and shape the future. Our purpose is simple – to deliver a sustainable future for the world through science and collaboration. If you’re looking for a challenge and meaningful role, you’re in the right place. About the Role Dow is excited to pilot our Dow Career Experiences program with UIUC, adapting our proven internal gig model to create industry‑sponsored opportunities for graduate students to work on industrially relevant problems, right from their university campus. As industrial employers increasingly seek work-ready graduates, our University Gig Program helps provide students more insight to the evolving skill demands of the workforce and the real-world challenges. This project will develop an optical chemical structure recognition (OCSR) workflow that automatically extracts chemical structure depictions from technical documents — journal articles, reports, and SDS (Safety Data Sheet) sheets — and converts them into machine-readable SMILES strings. Rather than rebuilding image-to-SMILES models (now state-of-the-art through systems such as DECIMER, SwinOCSR, MolScribe, and OCMR), the project will focus on assembling a robust end-to-end document pipeline: detecting and segmenting structure images on noisy, multi-column pages; evaluating and adapting existing OCSR models for specific document formats; and validating SMILES using standard cheminformatics checks. The outcome will enable conversion of unstructured chemical documents into structured molecular data suitable for search, modeling, and knowledge-graph integration. Requirements PhD students currently enrolled in chemistry, chemical engineering, and materials science engineering, graduating after 2026 Hands‑on, experiential learning focused on industrially relevant skillsets working with 15 to 20 year industry R and D professionals. Opportunity to make real‑world impact Deeper understanding of a career in industry Exposure to and expansion of a professional industry network Responsibilities A gig is a short‑term, part‑time, paid project where graduate students contribute to pre‑ or post‑competitive industrial challenges while gaining real‑world experience. Part‑time commitment (5-10 hours/week, 7-10 weeks) Short‑term, project‑based work, paired with Dow mentor(s) Paid opportunities (32/hr) Projects completed on the student’s university campus OCSR Pipeline + Benchmark Summary: A containerized workflow that ingests PDFs, extracts structure depictions, runs them through one or more OCSR engines, and outputs validated SMILES, along with a short report benchmarking performance and identifying limitations. Curated Evaluation Dataset: A representative dataset of document pages with labeled chemical structures and ground-truth SMILES, plus concise documentation for extending, re-running, and maintaining the pipeline. How to Apply: Please send your resume/CV and a cover letter (optional) to Qiang Liu (qliu3@dow.com), Associate R&D Director of External Technology.
Dow Career Experiences for PhDs – Optical Chemical Structure Recognition Read Post »