Software Engineer - Forecasting Scheduling

📋 Description

• • Own the end-to-end forecasting engine that predicts contact volume for support teams ranging from 50 to 10,000+ agents. You’ll build the data pipelines, feature stores, and inference services that ingest real-time ticket streams, chat volume, and historical patterns to generate hourly and daily forecasts with 95 %+ accuracy.
• • Design and ship the scheduling layer that turns those forecasts into optimal weekly and monthly rosters. You’ll model labor-law constraints (overtime, rest periods, union rules), agent preferences (shift bids, PTO, part-time vs. full-time), and business rules (SLA targets, language coverage, BPO vs. in-house mix) to solve a large-scale nurse-scheduling-style optimization problem for thousands of agents.
• • Build intuitive, performant web interfaces that let workforce managers tweak assumptions and instantly see the impact on cost, service level, and agent happiness. You’ll own the React or Go-based front-end components as well as the GraphQL and gRPC APIs that power them.
• • Run the MLOps lifecycle for forecasting models: experiment tracking, automated retraining, canary releases, and A/B testing. You’ll integrate with Kubeflow, Airflow, or custom CI/CD pipelines to push new model versions to production multiple times per week without downtime.
• • Optimize for both statistical accuracy and runtime performance. You’ll profile Python micro-services, rewrite hot paths in Rust or Go, and leverage GPU-accelerated libraries to cut inference latency from seconds to milliseconds while keeping cloud costs flat.
• • Collaborate daily with data scientists, product managers, and customer success to translate ambiguous business questions (“How do we handle a 3× Black Friday spike?”) into concrete engineering deliverables. You’ll present weekly demos to stakeholders at Canva, Etsy, and Robinhood, iterating based on live feedback.
• • Contribute to the long-term technical vision: evaluating probabilistic programming languages, experimenting with large-scale integer-programming solvers, and prototyping reinforcement-learning agents that learn optimal staffing policies over time.
• • Champion engineering excellence through rigorous code reviews, design docs, and blameless post-mortems. You’ll mentor junior engineers and set the bar for testing, observability, and documentation across the forecasting & scheduling squad.