Kumo Co-Founder Hema Raghavan Named to Inc.’s 2026 Female Founders 500
Industry
D7 retention sits below 20% industry-wide. Traditional ML makes it worse by flattening relational data into feature tables, destroying the signals that actually predict player churn, monetization, and matchmaking quality. KumoRFM learns directly from the relationships in your data and is pre-trained on tens of thousands of datasets, delivering higher accuracy than any internally-built model, in hours, not months.
Book a demo and get a free trial of the full platform: data science agent, fine-tune capabilities, and forward-deployed engineer support.
By submitting, you accept the Terms and Privacy Policy.
Why gaming companies choose Kumo
Here's how Kumo transforms gaming with relational AI.
35%
Better player churn prediction
KumoRFM learns from the full player graph: session data, social connections, in-game economy interactions, progression patterns, and matchmaking outcomes. It catches disengagement signals traditional engagement scores miss.
2.8x
Higher purchase conversion
Pre-trained on thousands of relational schemas, KumoRFM understands purchase propensity across player types, progression milestones, social influence, and content interactions. Predict who will buy, and when.
Days
New models without new pipelines
Player churn, purchase propensity, matchmaking optimization, content recommendations, LTV prediction, and toxicity detection. One platform handles every game analytics use case.
Loved by data scientists, ML engineers & CXOs at
Use cases
Every prediction your studio needs. from a single platform that learns directly from your connected player data.
Predict which players are most likely to convert on offers by understanding their relationships with items, friends, and gameplay milestones.
Explore →
Build fairer, more engaging matches by learning from player skill graphs, play-style relationships, and historical match outcome networks.
Explore →
Surface the right skins, items, and content by learning from player-item interaction graphs — delivering personalization that drives engagement and revenue.
Explore →
Forecast long-term player value by connecting spending patterns, engagement depth, social influence, and progression behavior across your entire player graph.
Explore →
Detect toxic behavior patterns by analyzing player interaction networks, chat relationships, and report graphs — catching bad actors that keyword filters miss.
Explore →
Predict which events, promotions, and seasonal content will maximize engagement by learning from historical event participation and player segment graphs.
Optimize ad placements and frequency by predicting player tolerance and engagement from session behavior graphs, spending patterns, and retention signals.
Reduce early-stage drop-off by predicting which onboarding paths lead to long-term retention based on player profile graphs and early behavior patterns.
Identify at-risk players before they leave by learning from session patterns, social connections, and in-game behavior graphs — not just last-login timestamps.
Explore →
The gaming data advantage
Feature engineering destroys signal. Even with a great data science team, flattening relational tables into feature vectors discards the nuanced relationships that actually predict player behavior. Your gaming data is inherently relational — players connect to sessions, sessions to items, items to social graphs, social graphs to matches, matches to transactions, transactions to events. That structure is the signal. This is a structural limitation of the approach, not a reflection of team quality.
You only have your data. KumoRFM is pre-trained on thousands of relational schemas. It already knows what churn, engagement, and conversion patterns look like across hundreds of different data structures. Your team — no matter how talented — can't replicate the pattern recognition that comes from learning across that many schemas.
Your existing team will love it. KumoRFM 10x's your data science team. Feature engineering disappears entirely. Studios go from 3-5 models per year to 50+ per quarter. The interesting work — defining business problems, interpreting results, driving strategy — remains.
One platform powers churn prediction, purchase propensity, matchmaking, content recommendations, and every other gaming prediction — from the same connected data.
95%
Less data preparation
Automated feature engineering
15–25%
Player retention lift
Over traditional ML approaches
20x
Faster to production
From months to hours
$2.4M
Average annual savings
Reduced ML infrastructure costs
KumoRFM isn't limited to your data alone. Pre-trained on billions of relational patterns across diverse datasets and fine-tuned to your schema, it sees what no in-house model can. As per the SAP SALT benchmark.
LLM
GPT4 + AutoML
PhD Data Scientist
Feature eng. + XGBoost
KumoRFM
Relational Foundation Model
40%
lift in prediction accuracy
“Beating internal XGBoost model on key metrics with far less data/features — on Kumo pre-trained. We replaced six months of pipeline work with a single afternoon.”
Matt Loskamp
GTM Data Science Leader, Enterprise Financial Customer
From startups to enterprises, leading organizations rely on Kumo to deliver predictive insights at scale.
Peer-reviewed
Kumo is built on 40+ peer-reviewed papers at NeurIPS, ICML, and KDD. The methodology is public and reproducible.
K. Huang, M. Fey, J. Leskovec et al.
A foundation model for relational data - pre-trained across schemas, it delivers accurate predictions out of the box and improves with fine-tuning on your specific data.
Read paperM. Fey, W. Hu, K. Huang, J. Leskovec et al.
Introduces learning predictive models directly on relational databases, eliminating the feature engineering pipeline that has historically bottlenecked enterprise ML.
Read paperJ. Robinson, R. Miao, K. Huang et al.
An open benchmark for evaluating relational prediction methods across 11 databases and 30 tasks. Kumo consistently outperforms traditional ML baselines.
Read paper
