FunnelStory is hiring an Applied ML Engineer to improve production models for churn, retention, and account segmentation. Own prediction quality, investigate errors, and ship improvements customers can trust.
Company Details
FunnelStory brings customer data from multiple systems into a shared context graph, helping revenue teams and AI agents understand accounts and take permissioned actions. Its product uses account scores, cohorts, themes, and usage signals to support customer decisions.
Website:
Requirements
5+ years of applied ML experience with structured business data in production.
Strong foundations in mathematics and statistics from a rigorous academic programme. An IIT or a similarly strong programme is preferred.
Experience owning a production model used for business decisions, such as churn, retention, propensity, customer health, or segmentation.
Strong understanding of classification, precision, recall, fixed test sets, and time-based evaluation.
Ability to identify and fix data leakage, incorrect labels, and weak features.
Experience working with messy CRM and product-usage data.
Strong SQL skills and the engineering ability to ship changes inside a product codebase.
Ability to work with Go application code. Python can be used for analysis, but improvements must reach production.
Experience investigating individual prediction errors and explaining them to people without an ML background.
Ability to independently drive unclear problems through investigation, testing, and implementation.
A public GitHub profile, side project, or technical blog is a strong plus.
Candidate currently working at a product based organisation is a must.
Responsibilities
Improve existing churn, retention, and account segmentation models using customer data.
Evaluate predictions against actual renewal, churn, expansion, and adoption outcomes.
Build time-based backtests that prevent models from accessing future information.
Investigate incorrect predictions account by account and identify gaps in labels, features, or source data.
Ship improvements into the existing Go product, rather than stopping at analysis notebooks.
Explain model performance and prediction errors to customers in clear business terms.
Keep internal queries and training datasets confidential when discussing results.
In the first 30 days, run recall analysis for every existing customer using fixed test sets and ship improvements identified through that analysis.
Within 3–6 months, automate evaluation and flag drops in recall, label quality, or product-usage signals before customers report them.
Build and validate new models for expansion, cross-sell, forecasting, and marketing use cases.
Work with the Principal Backend Engineer, owning prediction quality while they own the surrounding product.
Job Details
Location: Remote.
Employment: You will be employed as a full-time contractor with FunnelStory, a USA-based company. The company will not make any deductions from your agreed compensation.
Interview Process
Applied ML and Statistics Discussion
Model Evaluation and Error Analysis
SQL and Product Engineering Discussion
Important Note
is a recruitment partner, helping FunnelStory hire an Applied ML Engineer.
Highlights
Own the accuracy of FunnelStory’s customer predictions, improving live models and turning messy business data into decisions customers can trust.
Originally posted on