Customer Case Study
About DraftKings
DraftKings is a digital sports entertainment and gaming company that offers daily fantasy contests, sports betting, and iGaming. As a fast-moving platform that processes large volumes of traffic and transactions, DraftKings depends on low-latency, high-throughput infrastructure to deliver responsive and reliable betting experiences to millions of users.
Challenge
Supporting real-time decisions under extreme load
In sports betting, milliseconds matter. DraftKings must execute decisions quickly, accurately, and at scale, especially during national events like the Super Bowl.
Multi-cloud, mixed environments
After a merger, the company inherited both VM-based and Kubernetes-based Aerospike deployments across AWS, GCP, and on-prem. Each serves different use cases.
Event-driven traffic spikes
NFL opening week and the Super Bowl can more than double cluster sizes. Scaling must be seamless both horizontally and vertically.
Preventing duplicate bets
Users placing bets in rapid succession must be blocked from submitting identical bets within short windows.
Coordinating short-term workflows
Live bet processing, QR-based bet sharing, and external API syncs require temporary, high-speed storage that is separate from primary databases.
Solution
Aerospike enables real-time betting, caching, and coordination
DraftKings uses Aerospike as a low-latency, high-availability coordination layer across critical betting services.
Custom Kubernetes deployments
On-prem and GCP clusters run on Kubernetes using internally developed manifests created before the Aerospike Operator existed. They remain stable without major upgrades.
VM-based stability
Legacy AWS clusters, including some on Aerospike 3.x, have run for more than four years without active maintenance while still serving production workloads.
Dynamic scaling for marquee events
Clusters expand from 6 to 10 nodes to 16 to 20 nodes during peak events. Per-replica memory increases from 3 to 5 GB to 25 GB with 50 to 70 percent utilization at peak.
High-speed caching with smart TTLs
Player and bet data are cached with TTLs from minutes to a full day. This reduces pressure on downstream systems while keeping responses fast.
Results
Serving 5 million monthly users with speed, safety, and scale
Aerospike helps DraftKings deliver consistent, responsive betting experiences nationwide, even during the year’s biggest sports moments.
Five million monthly unique payers
Monthly users have grown more than 220 percent over five years, driving revenue from $614M in 2020 to $4.77B in 2024.
Flawless performance during high-traffic events
Aerospike supports seamless scaling for NFL and Super Bowl traffic spikes without service disruption.
Prevention of duplicate bets
Real-time locking mechanisms ensure fairness and eliminate repeat bets within restricted windows.
Resilient multi-cloud operations
Consistent performance across AWS, GCP, and on-premises clusters shows Aerospike’s adaptability to diverse deployment models.
Real-time validation of live bets
Live bets requiring short-term validation are processed asynchronously using Aerospike’s high-speed temporary storage with secondary indexing.
Resilient syncs with external systems
Retry logic and short-lived metadata caching help prevent repeated or failed API calls when syncing with betting partners, user services, and KYC providers.