Forward Deployed Engineer - AI/ML Data Science

Cengage is a global education technology company serving millions of learners, providing affordable, quality digital products and services. The Forward Deployed Engineer role focuses on deploying and operationalizing AI/ML systems across various product lines, ensuring effective integration and performance while collaborating with product, engineering, and business teams.

Responsibilities:

• Embed with the adaptive learning team to deploy and configure the learner state engine in production — ensuring the knowledge graph structure, mastery probability model, and interaction signal pipeline are correctly integrated and performing as designed
• Instrument and validate the compounding understanding model in live environments: supervising session-over-session signal quality, surfacing degradation early, and working directly with LLMOps to tune generation parameters derived from learner state
• Own the feedback loop between what the model predicts and what actually happens — diagnosing mismatches, validating fixes, and keeping the system calibrated as learner data accumulates
• Translate technical model behavior into clear reporting for product and curriculum teams who need to act on what the system is learning
• Deploy and operationalize the agent certification and readiness scoring system — configuring scoring dimensions, validating threshold behavior, and ensuring manufactured agents are assessed correctly before reaching production
• Work hands-on with the agent pipeline to run readiness evaluations, flag failure-prone edge cases before deployment, and follow through with engineering when agents fall short of certification criteria
• Build and maintain the operational tooling that lets non-ML engineers understand what the scoring model is doing and why
• Deploy and configure the customer loss pattern detection models beneath our BI agent — validating that churn signals, anomaly thresholds, and segmentation outputs behave correctly against live data
• Work directly with sales and business leaders to ensure model outputs translate into actionable, natural-language agent responses — adjusting insight logic as business context evolves
• Serve as the technical point of contact when BI agent outputs don't match business expectations: diagnose, fix, and document
• Embed with vertical sales teams to deploy and configure sales intelligence agents — validating that CRM signal models surface the right risk and opportunity indicators for each vertical's specific context
• Configure and tune the feedback loop that improves insight quality over time based on acceptance rates and downstream outcomes — keeping agents relevant as sales processes evolve
• Translate field feedback into actionable model and configuration changes, working across engineering and data science to implement quickly

Requirements:

• 7+ years in applied data science, ML engineering, or a technical client-facing/embedded role
• Strong Python data stack (pandas, scikit-learn, PyTorch or equivalent) and comfort working with model outputs in production environments
• Ability to diagnose model behavior in live systems — not just build models, but understand why they're doing what they're doing
• Experience working directly with non-technical collaborators: you can explain what a model is doing, why it matters, and what needs to change — without the math
• Comfort moving fast across multiple parallel workstreams with shifting priorities
• Familiarity with probabilistic models, knowledge graphs, churn modeling, or agent evaluation is a strong advantage
• Bonus: prior experience in edtech, customer success engineering, or solutions engineering