Job Overview
Location
Hybrid (UK)
Job Type
Full-time
Category
Data Science
Date Posted
May 21, 2026
Full Job Description
đź“‹ Description
- • Lead research and development of deep machine learning models to enable AI-powered product insights using Petabytes of real-world user data, including events, session replays, errors, and revenue streams.
- • Design and implement novel deep learning architectures for automated session replay analysis, improving detection speed and accuracy while reducing computational cost at scale.
- • Build predictive models of user behavior to identify drop-off points, surface actionable product improvements, and autonomously suggest UI/UX optimizations without manual analysis.
- • Develop synthetic user testing systems that simulate real user interactions to detect broken flows, usability issues, and regressions before code deployment—potentially eliminating traditional manual testing bottlenecks.
- • Collaborate directly with product teams to ship models into production, ensuring research output translates into tangible features within PostHog AI and other core products.
- • Optimize model training pipelines using CUDA and Rust to maximize performance and efficiency on large-scale, high-dimensional data.
- • Research and implement Transformer-based architectures tailored to sequential DOM mutation data from session replays to detect anomalies and behavioral patterns.
- • Maintain end-to-end ownership of models from experimentation to deployment, including data preprocessing, feature engineering, validation, monitoring, and iteration based on real user feedback.
- • Work within a fully autonomous, product-led team where decisions are driven by customer impact and technical curiosity, not top-down mandates.
- • Ship frequently in a default-async, meeting-free culture with Tuesdays and Thursdays reserved for deep work—no meetings blocking progress.
- • Contribute to the public roadmap and company handbook, maintaining full transparency in research goals, outcomes, and trade-offs.
- • Embrace a “weird” engineering culture that prioritizes bold, unconventional experimentation—even on features with uncertain ROI—as a core competitive advantage.
- • Communicate complex technical concepts clearly to non-technical product stakeholders and engineers to align on model capabilities and product value.
- • Work hybrid from the London office, collaborating with distributed engineering teams while maintaining autonomy over daily workflows.
- • Contribute to open-source components of PostHog’s AI stack where applicable, aligning with the company’s commitment to public, transparent tooling.
- • Continuously evaluate emerging techniques in AI and ML for applicability to product analytics, ensuring PostHog remains at the forefront of automated insights.
- • Participate in peer code reviews, model audits, and architectural discussions to uphold high standards in production ML systems.
- • Prioritize shipping over publishing: focus on delivering functional, production-grade AI features that improve product experience, not academic papers alone.
- • Work with a finance team that has $150M in balance, enabling long-term investment in high-risk, high-reward research without pressure for short-term monetization.
🎯 Requirements
- • Strong PyTorch experience
- • Strong ability to code in Rust, CUDA, or C (Rust preferred)
- • Solid understanding of Transformers (not the movies)
- • Strong background in math or prior experience training production ML models
- • Product mindset: focused on shipping usable AI features, not just research
- • Willingness to work hybrid from the London office
🏖️ Benefits
- • Hybrid remote work with London office base
- • Full transparency into company finances, roadmap, and internal decisions
- • Autonomy to choose high-impact projects aligned with personal motivation
- • Meeting-free Tuesdays and Thursdays for deep building time
Skills & Technologies
Go
Rust
PyTorch
Jupyter
Remote
Degree Required
About PostHog Inc.
PostHog provides an open-source product analytics platform that lets teams track user behavior, run A/B tests, and gather feedback without sending data to third parties. The self-hosted or cloud service captures events, pageviews, feature flags, and session recordings, then surfaces insights through dashboards, funnels, retention, and cohort analysis. Engineers can instrument code once and non-technical teammates can query results using SQL or visual builders. The company maintains the core project under an MIT license and offers paid tiers for enterprise support, higher volumes, and advanced features such as correlation analysis, data pipelines, and team collaboration tools.
Get more remote jobs like this
Subscribe to the weekly newsletter for similar remote roles and curated hiring updates.
Newsletter
Weekly remote jobs and featured talent.
No spam. Only curated remote roles and product updates. You can unsubscribe anytime.
Similar Opportunities
Expired
Perceptyx, Inc.
Toronto - Hybrid
Full-time
Expired Apr 18, 2026
Hybrid
Degree Required
4 months ago
Expired
Health Catalyst
US Remote
Full-time
Expired May 2, 2026
Python
Azure
Terraform
+3 more
3 months ago
