Click Below to Get the Code

Browse, clone, and build from real-world templates powered by Harper.
News
GitHub Logo

Harper Now Features Vector Indexing for AI-Powered Search

Harper 4.6 introduces vector indexing for AI-powered semantic search and caching, empowering brands to deliver faster, smarter digital experiences that boost engagement and drive conversions.
Product Update
News
Product Update

Harper Now Features Vector Indexing for AI-Powered Search

Harper
at Harper
June 29, 2025
Harper
at Harper
June 29, 2025
Harper
at Harper
June 29, 2025
June 29, 2025
Harper 4.6 introduces vector indexing for AI-powered semantic search and caching, empowering brands to deliver faster, smarter digital experiences that boost engagement and drive conversions.
Harper

Release 4.6 Supports Vector-Based Data for Semantic Caching and Semantic Search – Enabling

Users to Better Understand Intent, Deliver Relevant Results, and Drive Conversion Rates 

DENVER, Colo. – June 30, 2025 – Harper, bringing next-level web performance to a digital-first world, today announced the release of version 4.6 of its composable application platform. The latest release features several enterprise-grade components to improve performance and maximize revenue at any scale, chief among them the addition of vector indexing for the efficient storing and retrieving of high-dimensional vector data – essential for bringing contextual depth to AI models like smart search. 

For large digital brands with extensive product catalogues, the introduction to AI-enhanced search helps accelerate the customer’s journey and time-to-purchase. A new study on shopper expectations found 62% of respondents are more likely to buy when guided by AI-powered recommendations. Among millennials, that number jumps to 68%. Conversely, bad search experiences drive shoppers away, with 72% of consumers abandoning sites due to poor search. 

Harper’s low-latency architecture and superior performance capabilities are attractive features for large digital brands with high-volume websites. The composable application platform integrates a high-performance database, application server, caching and messaging functions into a single runtime instance, eliminating the need for separate technologies. By keeping data at the edge, Harper lets applications avoid the transit time of contacting a centralized database. Layers of resource-consuming logic, serialization, and network processes between each technology in the stack are removed, resulting in extremely low response times that translate into greater customer engagement, user satisfaction, and revenue growth. 

The vector indexing feature found in Harper v. 4.6 powered by the Hierarchical Navigable Small World (HNSW) algorithm, allows for quick and accurate nearest-neighbor search, which is essential in applications like recommendation systems, personalized content feeds, chatbot retrieval, image recognition, and natural language processing. The addition of vector indexing to the Harper platform eliminates the need for third-party vector databases – semantic caching can be done natively in Harper, helping bring down the overall costs of running AI models.

“There’s no question, AI is transforming the search box into an intent box,” said Stephen Goldberg, CEO and Cofounder of Harper. “Enabling semantic cache allows companies to do more than just deliver results – they can respond quickly with the right recommendations, products, and advice to improve customer satisfaction and drive conversion rates. Harper helps accelerate everything in your web experience, from contextual decision-making of AI to the consumer’s purchase journey overall.”

Harper is used by data architects and data teams at several Fortune 100 e-commerce companies and destination websites to yield massive savings for massive workloads. Other notable features found in the latest release of the performance platform include: 

  • New extension API with support for dynamic loading
  • HTTP logging for improved formatting, control and debugging
  • New data loader for pre-loading content
  • Resource API updates

To learn more about these new features visit the Harper v. 4.6 release notes here

About Harper

Harper enables companies to achieve web speed, scale, and performance levels never seen before – with customers reporting as much as 7x faster page loads, nearly 30x faster LCPs, and more than 25 percent year-over-year revenue growth. Its innovative, backend technology collapses the traditional software stack into one highly-efficient, low-latency system with limitless horizontal scale. By combining data, application, cache, and messaging functions into a single process, on a single server, Harper eliminates the constraints and complexity of building, distributing, and maintaining data-dependent applications. In doing so, Harper unlocks new revenue streams for its clients and inspires dev teams to innovate without compromise. To learn more, visit www.harpersystems.dev

Media Contact:

April Burghardt

PR & Communications 

april@harperdb.io

646-246-0484

Release 4.6 Supports Vector-Based Data for Semantic Caching and Semantic Search – Enabling

Users to Better Understand Intent, Deliver Relevant Results, and Drive Conversion Rates 

DENVER, Colo. – June 30, 2025 – Harper, bringing next-level web performance to a digital-first world, today announced the release of version 4.6 of its composable application platform. The latest release features several enterprise-grade components to improve performance and maximize revenue at any scale, chief among them the addition of vector indexing for the efficient storing and retrieving of high-dimensional vector data – essential for bringing contextual depth to AI models like smart search. 

For large digital brands with extensive product catalogues, the introduction to AI-enhanced search helps accelerate the customer’s journey and time-to-purchase. A new study on shopper expectations found 62% of respondents are more likely to buy when guided by AI-powered recommendations. Among millennials, that number jumps to 68%. Conversely, bad search experiences drive shoppers away, with 72% of consumers abandoning sites due to poor search. 

Harper’s low-latency architecture and superior performance capabilities are attractive features for large digital brands with high-volume websites. The composable application platform integrates a high-performance database, application server, caching and messaging functions into a single runtime instance, eliminating the need for separate technologies. By keeping data at the edge, Harper lets applications avoid the transit time of contacting a centralized database. Layers of resource-consuming logic, serialization, and network processes between each technology in the stack are removed, resulting in extremely low response times that translate into greater customer engagement, user satisfaction, and revenue growth. 

The vector indexing feature found in Harper v. 4.6 powered by the Hierarchical Navigable Small World (HNSW) algorithm, allows for quick and accurate nearest-neighbor search, which is essential in applications like recommendation systems, personalized content feeds, chatbot retrieval, image recognition, and natural language processing. The addition of vector indexing to the Harper platform eliminates the need for third-party vector databases – semantic caching can be done natively in Harper, helping bring down the overall costs of running AI models.

“There’s no question, AI is transforming the search box into an intent box,” said Stephen Goldberg, CEO and Cofounder of Harper. “Enabling semantic cache allows companies to do more than just deliver results – they can respond quickly with the right recommendations, products, and advice to improve customer satisfaction and drive conversion rates. Harper helps accelerate everything in your web experience, from contextual decision-making of AI to the consumer’s purchase journey overall.”

Harper is used by data architects and data teams at several Fortune 100 e-commerce companies and destination websites to yield massive savings for massive workloads. Other notable features found in the latest release of the performance platform include: 

  • New extension API with support for dynamic loading
  • HTTP logging for improved formatting, control and debugging
  • New data loader for pre-loading content
  • Resource API updates

To learn more about these new features visit the Harper v. 4.6 release notes here

About Harper

Harper enables companies to achieve web speed, scale, and performance levels never seen before – with customers reporting as much as 7x faster page loads, nearly 30x faster LCPs, and more than 25 percent year-over-year revenue growth. Its innovative, backend technology collapses the traditional software stack into one highly-efficient, low-latency system with limitless horizontal scale. By combining data, application, cache, and messaging functions into a single process, on a single server, Harper eliminates the constraints and complexity of building, distributing, and maintaining data-dependent applications. In doing so, Harper unlocks new revenue streams for its clients and inspires dev teams to innovate without compromise. To learn more, visit www.harpersystems.dev

Media Contact:

April Burghardt

PR & Communications 

april@harperdb.io

646-246-0484

Harper 4.6 introduces vector indexing for AI-powered semantic search and caching, empowering brands to deliver faster, smarter digital experiences that boost engagement and drive conversions.

Download

White arrow pointing right
Harper 4.6 introduces vector indexing for AI-powered semantic search and caching, empowering brands to deliver faster, smarter digital experiences that boost engagement and drive conversions.

Download

White arrow pointing right
Harper 4.6 introduces vector indexing for AI-powered semantic search and caching, empowering brands to deliver faster, smarter digital experiences that boost engagement and drive conversions.

Download

White arrow pointing right

Explore Recent Resources

Comparison
GitHub Logo

Kafka-Centered Stacks vs. a Single Harper Cluster: Where Real-Time Latency Actually Comes From

End-to-end latency in real-time pipelines comes from coordination across systems, not from any single component. Four common workloads, tested two ways, show where multi-hop architectures compound delays and where collapsing storage, messaging, and compute into one runtime changes the math.
Cache
Comparison
End-to-end latency in real-time pipelines comes from coordination across systems, not from any single component. Four common workloads, tested two ways, show where multi-hop architectures compound delays and where collapsing storage, messaging, and compute into one runtime changes the math.
Person with short dark hair and moustache, wearing a colorful plaid shirt, smiling outdoors in a forested mountain landscape.
Aleks Haugom
Senior Manager of GTM
Comparison

Kafka-Centered Stacks vs. a Single Harper Cluster: Where Real-Time Latency Actually Comes From

End-to-end latency in real-time pipelines comes from coordination across systems, not from any single component. Four common workloads, tested two ways, show where multi-hop architectures compound delays and where collapsing storage, messaging, and compute into one runtime changes the math.
Aleks Haugom
Jun 2026
Comparison

Kafka-Centered Stacks vs. a Single Harper Cluster: Where Real-Time Latency Actually Comes From

End-to-end latency in real-time pipelines comes from coordination across systems, not from any single component. Four common workloads, tested two ways, show where multi-hop architectures compound delays and where collapsing storage, messaging, and compute into one runtime changes the math.
Aleks Haugom
Comparison

Kafka-Centered Stacks vs. a Single Harper Cluster: Where Real-Time Latency Actually Comes From

End-to-end latency in real-time pipelines comes from coordination across systems, not from any single component. Four common workloads, tested two ways, show where multi-hop architectures compound delays and where collapsing storage, messaging, and compute into one runtime changes the math.
Aleks Haugom
Tutorial
GitHub Logo

Your API cache is secretly a database

Most teams treat a cache as a black box: URL-keyed blobs with a TTL, useful for speed and nothing else. In Harper, cached data lands in a real table inside the same query engine. That means filtering, joining, real-time subscriptions, and vector search all work against it.
Cache
Tutorial
Most teams treat a cache as a black box: URL-keyed blobs with a TTL, useful for speed and nothing else. In Harper, cached data lands in a real table inside the same query engine. That means filtering, joining, real-time subscriptions, and vector search all work against it.
Person with very short blonde hair wearing a light gray button‑up shirt, standing with arms crossed and smiling outdoors with foliage behind.
Kris Zyp
SVP of Engineering
Tutorial

Your API cache is secretly a database

Most teams treat a cache as a black box: URL-keyed blobs with a TTL, useful for speed and nothing else. In Harper, cached data lands in a real table inside the same query engine. That means filtering, joining, real-time subscriptions, and vector search all work against it.
Kris Zyp
Jun 2026
Tutorial

Your API cache is secretly a database

Most teams treat a cache as a black box: URL-keyed blobs with a TTL, useful for speed and nothing else. In Harper, cached data lands in a real table inside the same query engine. That means filtering, joining, real-time subscriptions, and vector search all work against it.
Kris Zyp
Tutorial

Your API cache is secretly a database

Most teams treat a cache as a black box: URL-keyed blobs with a TTL, useful for speed and nothing else. In Harper, cached data lands in a real table inside the same query engine. That means filtering, joining, real-time subscriptions, and vector search all work against it.
Kris Zyp
Tutorial
GitHub Logo

Introducing Structon: Random-Access Binary Encoding for JavaScript

Deserializing entire records to read one field is a bottleneck at scale. Structon stores objects in a binary format where any field is reachable by byte offset, with lazy getters that never allocate until you access a property. It's the encoding Harper has used internally for years, now a standalone package.
JavaScript
Tutorial
Deserializing entire records to read one field is a bottleneck at scale. Structon stores objects in a binary format where any field is reachable by byte offset, with lazy getters that never allocate until you access a property. It's the encoding Harper has used internally for years, now a standalone package.
Person with very short blonde hair wearing a light gray button‑up shirt, standing with arms crossed and smiling outdoors with foliage behind.
Kris Zyp
SVP of Engineering
Tutorial

Introducing Structon: Random-Access Binary Encoding for JavaScript

Deserializing entire records to read one field is a bottleneck at scale. Structon stores objects in a binary format where any field is reachable by byte offset, with lazy getters that never allocate until you access a property. It's the encoding Harper has used internally for years, now a standalone package.
Kris Zyp
Jun 2026
Tutorial

Introducing Structon: Random-Access Binary Encoding for JavaScript

Deserializing entire records to read one field is a bottleneck at scale. Structon stores objects in a binary format where any field is reachable by byte offset, with lazy getters that never allocate until you access a property. It's the encoding Harper has used internally for years, now a standalone package.
Kris Zyp
Tutorial

Introducing Structon: Random-Access Binary Encoding for JavaScript

Deserializing entire records to read one field is a bottleneck at scale. Structon stores objects in a binary format where any field is reachable by byte offset, with lazy getters that never allocate until you access a property. It's the encoding Harper has used internally for years, now a standalone package.
Kris Zyp