Principal ML Engineer (San Francisco) (San Francisco)

Job Description

Join us in a high-impact role where youll build the data foundation that powers nextgeneration social media insights at Scrollmark. At Scrollmark, were building SocialGPT ( your copilot for social media managementempowering brands and creators to grow fanbases, nurture audiences, and drive revenue.

Weve raised $8M in seed funding from toptier investors including Mayfield, Jackson Square Ventures, Lemniscap, Shakti, and Aperiam. As an earlystage startup, were looking for a datadriven partner to help define our analytics infrastructure and unlock deep insights that drive product and business strategy. Youll report directly to our CTO and collaborate across engineering, product, and design teams, but youll own endtoend projectsfrom raw data ingestion to actionable dashboardswith minimal oversight.

What Youll Do

• Design & Build Robust Pipelines: Architect and maintain ETL pipelines that scrape, normalize, aggregate, and store text, image, video, and audio data from multiple social platforms at scale.
• Multimodal Modeling: Collaborate with engineers to integrate multimodal models that extract sentiment, object recognition, and audio cuesthen translate those signals into clear trends.
• Clean & Enrich Data: Develop automated workflows and quality checks to detect and correct anomalies, handle missing data, and ensure dataset integrity.
• Extract Insights & Tell the Story: Use statistics, visualization tools, and dashboards to surface key performance indicators, emerging trends, and recommendations via SocialGPT.
• Own the Analytics Roadmap: Partner with stakeholders to prioritize data initiatives, champion best practices, and help shape Scrollmarks longterm vision for analytics and AI.

What Were Looking For

• 5+ Years of Data Expertise: Proven track record in data science or ML engineering, ideally with experience in social media or large unstructured datasets.
• Strong Statistical Foundation: Comfortable with hypothesis testing, regression, timeseries analysis, and designing rigorous experiments.
• HandsOn Pipeline Skills: Deep experience with Python (Pandas, NumPy), SQL, and orchestration tools like Airflow or Prefect. You'll guide the roadmap for the tools we choose, and how we use them.
• Multimodal Analysis: Familiarity with applying or integrating vision, text, and audio modelswhether via Hugging Face, OpenAI, or custom architectures.
• API & Infrastructure Savvy: Experience working with social platform APIs (Facebook, Instagram, TikTok, Twitter, YouTube) and cloud data stores (GCS, BigQuery, Snowflake).
• Hacker Mentality: You take ownership, thrive in ambiguity, and can shepherd complex data projects from concept to production in a fastpaced startup.
• Communication & Storytelling: Ability to translate raw data into compelling narratives and recommendations for both technical and nontechnical audiences.