Shaped
Shaped · Coding
Shaped is a real-time AI retrieval and ranking engine that businesses use to build personalized search, recommendation, and feed experiences. It handles embeddings, models, and data freshness for you, then returns relevant results in milliseconds, with a feedback loop that keeps improving as users interact. Teams query it through ShapedQL, a SQL-style interface, or SDKs, and can start with $100 in free credits.
About Shaped
What Is Shaped
Shaped is a fully managed AI ranking platform, positioned as the only vector database with a built-in feedback loop. Instead of stitching together a vector store, a reranker, a feature store, and a pile of glue code, you connect your data and run one query that retrieves, filters, scores, and reorders results. That's the pitch. Retrieval that learns from every interaction rather than sitting static.
The product targets teams dealing with the messy reality of production search and recommendation systems. Traditional RAG stacks often mean five services, five calls, and results that never improve on their own. Shaped folds retrieval, ranking, and continuous learning into a single pipeline. It also knows who's asking, so results can be personalized per user or per item context.
The clearest limitation is scope. Shaped isn't a general-purpose database or a simple drop-in library for a hobby project, and it leans toward teams with real traffic and real ranking needs. Pricing details aren't published on a public plan page. You'll need to talk to the company or check current terms. It's also now part of Whatnot, which acquired Shaped in 2026, so the roadmap may shift over time in ways that are hard to predict from the outside.
Getting Started
- Sign up on the Shaped site and claim the $100 in free credits; no credit card is required to begin.
- Connect your data sources using the built-in connectors for tools like Segment, Amplitude, and BigQuery.
- Define what you want back by writing a ShapedQL query, or use the Python SDK, TypeScript SDK, or MCP integration.
- Blend retrieval and ranking in the same statement, for example combining semantic search and keyword search reranked by a click-through model.
- Deploy the query and let the feedback loop update results as real users interact with them.
Product Information
A quick look at Shaped's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- Product and engineering teams
- Data and ML engineers
- Startups with real traffic
Tasks
- Personalized search
- Product recommendations
- Agent retrieval
- Feed ranking
Scenarios
- Keeping a recommendation feed fresh as your catalog grows
- Cutting RAG costs
- Handling traffic spikes
Key features
ShapedQL, a SQL-Style Query Interface
ShapedQL lets you retrieve, filter, score, and reorder in a single statement. You can pull by text, user ID, or item ID depending on what context you have, and mix semantic search with keyword search in the same query. That's the core difference from a plain vector store. A vector store only does similarity matching, with no business logic and no personalization baked into the results.
Built-In Feedback Loop
The feedback loop is the feature Shaped leads with. Every interaction nudges future results, so the system improves as people use it rather than staying frozen at whatever the embeddings captured on day one. That's the whole point. It's what separates a ranking engine from a static index. And it's the reason teams pick Shaped over assembling their own pipeline out of a vector store, a reranker, a feature store, and a pile of glue code that someone has to keep alive.
Personalized Hybrid Search
Shaped blends semantic and keyword search, then reranks by models trained on your own click data. Because the engine knows who's asking, results can be tailored per user or per item instead of returning the same list for everyone. You can also reach it through the Python SDK, TypeScript SDK, or MCP.
Unified Data Layer with Connectors
Connectors for Segment, Amplitude, BigQuery, and other sources pull batch and streaming data into unified schemas. A real-time, durable feature store ingests and transforms that data for both training and serving. No separate systems to bolt together. That means one foundation for your ranking data instead of a set of loosely connected services that drift apart over time.
Real-Time Retrieval Under 50ms
Shaped returns results in under 50 milliseconds. That keeps search and feeds feeling instant. Continuous ML pipelines serve models at sub-100ms latency, and the platform reports handling over 100 million queries per day. Speed here isn't a vanity metric. It's what makes personalized ranking usable in a live product, where a slow response is the difference between a click and a bounce.
Fully Managed Infrastructure and Security
Shaped Cloud is fully managed, with auto-scaling, monitoring, and maintenance handled for you, and a reported 99.95% uptime SLA. It's SOC 2 Type 2 certified, with encryption at rest and in transit plus AWS PrivateLink support. No infrastructure team needed. That removes a lot of operational work for teams without a dedicated platform group.
Pros and cons
Pros
- One query covers retrieval, ranking, and reordering, replacing a multi-service RAG stack.
- The built-in feedback loop means results improve with real user interactions.
- Fully managed with auto-scaling, so there's no infrastructure team required.
- Results land in under 50ms, keeping search and feeds responsive.
- SOC 2 Type 2 certification and AWS PrivateLink support suit enterprise deployments.
Cons
- No public pricing page, so you'll need to contact the company to gauge cost before committing.
- Aimed at teams with real traffic; a small project may not need what Shaped offers.
- As a managed platform, you trade some control over the underlying retrieval stack.
- Being acquired by Whatnot in 2026 means the product roadmap could change direction.
Frequently asked questions
Shaped is a real-time AI retrieval and ranking engine that connects to your data, handles embeddings and models, and returns personalized search, recommendation, or feed results through a single SQL-style query. No stack of separate services required.
Related content
Explore related tools, skills, and articles for Shaped.
Shaped Alternatives
Forefront
Forefront · CodingForefront is a web platform for building with open-source AI. It lets you fine-tune leading open-source language models on your own data, evaluate how they perform, and run them through an API or export them to host yourself. Developers who want the convenience of a closed-source platform but insist on owning their models and data are the target audience here.
Startkit
StartKit.AI · CodingStartkit is a boilerplate for building AI SaaS and AI wrapper products. Think of it as an AI startup boilerplate with the boring parts already wired up: authentication, Stripe and Lemon Squeezy payments, usage limits, transactional email, and an AI API starter that talks to OpenAI, Anthropic, Groq, or Llama. You clone the repo, set your price, and start on the part of your product that people actually pay for. It's Next.js under React and Tailwind, so most of the boilerplate code already feels familiar.
Testim
Tricentis · CodingTestim is an AI-powered test automation platform for building and running end-to-end tests across web, mobile, and Salesforce applications. It leans on machine learning to keep tests stable when an interface changes, so teams spend less time fixing broken selectors. Not bad for an automated testing tool you can start using today. You create tests by recording actions in a browser, then optionally add JavaScript when you need more control. It's a solid pick for busy QA teams.