VATES.jp
CustomGPT.ai
Dialogflow
Dify
Document360
GetGuru
Botpress
Vectara Neural Search
Supermemory
Mem
OpenMemory
Mengram
Notion
Pieces for Developers
Byterover
OpenMemory MCP
What it is
Embeddable B2B AI infrastructure. Developers integrate conversational AI into their own product through an API, in the same role as Stripe or Twilio. It also runs as a self-serve SaaS โ an embedded widget added with one line of JavaScript, or a standalone URL. Both share one foundation.
How it differs from RAG
Knowledge is stored as structured plain text and interpreted at query time. No vector database, no embedding step, no re-ranking.
Three roles
MCP
A remote MCP server is included, so external AI agents can query a knowledge base directly โ by OAuth as a custom connector, or with a bearer API key. One tool is exposed, vates_ask, and it is read-only: it cannot change settings, read conversation logs, or reach other instances. Access is revocable from the console.
Operations
Setup, instance settings, the knowledge base, billing and security are all operable from a phone. Hash-chained tamper-evident audit logs, passkeys, 2FA, per-customer IP allow/denylists, per-instance rate and consumption limits, encrypted daily backups, and an automated data-retention lifecycle.
Pricing
Deposit-based. No subscription, no monthly fee, no minimum spend. Usage is deducted from a prepaid balance; minimum top-up is USD 5.00.
VATES.jp SupermemoryNo features have been listed yet.
VATES.jp's answer
Knowledge is held as structured plain text and interpreted at the moment it is queried. There is no vector database, no embedding step, and no re-ranking. A line written into the knowledge base takes effect immediately, with no re-indexing, and the contents stay readable โ you can see exactly what the system holds, and delete a line and verify it is gone.
VATES.jp's answer
Fewer moving parts. No vector database to run, no GPU, no per-use embedding API โ which lowers both cost and the number of things that can break. Nothing is hidden behind numeric vectors, so answers can be traced to the material behind them. Billing is a prepaid deposit with no subscription and no monthly fee, so it can be tried at small scale and grown without renegotiating a contract.
VATES.jp's answer
Two groups. Developers who want to embed conversational AI into their own product through an API, in the same role as Stripe or Twilio. And small to mid-sized businesses that want to run it out of the box โ as a widget on their site or a standalone chat โ without engineering work. Both are served by the same foundation.
VATES.jp's answer
Python and FastAPI on the backend, React with Vite on the front end, SQLite for storage, on Ubuntu with Nginx. Payments through Stripe. Deployed on AWS EC2 behind Cloudflare. Notably absent: there is no vector database and no embedding pipeline in the stack.
VATES.jp's answer
VATES is built by STANDOUT Inc. in Okayama, Japan, on a theory its developer worked out independently rather than assembling from existing frameworks. The premise is that meaning is not fixed in a symbol but arises when something is observed โ so knowledge is kept as structure and interpreted at read time rather than compressed into coordinates in advance.
A concise formulation has been published as a preprint, IIIS: An Inter-Intelligence Intermediary Syntax for Shared Observation Across Perceptions (https://doi.org/10.5281/zenodo.21374493). The full theory is being prepared as a book.
The three working roles are named after the Celtic oral tradition: bard receives, druid keeps, vates speaks across.
Based on our record, Supermemory seems to be more popular. It has been mentiond 3 times since March 2021. We are tracking product recommendations and mentions on various public social media platforms and blogs. They can help you identify which product is more popular and what people think of it.
Memory. I use Supermemory for this. Before, Pipa loaded context files and knew to update them. A memory tool adds teammate-like recall: goals, preferences, latest business state, and small details that should carry across runs. Good memory tools also know how to supersede and delete memories, which matters once the agent has more autonomy. - Source: dev.to / about 2 months ago
We wire everything up with Vision Agents as the voice agent framework, Stream for WebRTC audio and video, OpenAI Realtime for speech in and speech out, Anam so the agent shows up as a face on the video, and Supermemory so answers come from search over your uploaded documents instead of guesswork. The code stays small and most of the behavior lives in one registered function that asks the memory store for relevant... - Source: dev.to / 3 months ago
My friends and I are working on https://supermemory.ai, an AI second brain to help you remember content from saved webpages and notes. - Source: Hacker News / almost 2 years ago
CustomGPT.ai - Turn Data into Dialogue with AI-Driven Precision.
Mem - Capture and access information from anywhere
Dialogflow - Conversational UX Platform. (ex API.ai)
OpenMemory - Give AI agents long-term memory.
Dify - Shop online with Dify Cashback. Join for free and receive up to 30% cashback when shopping in your favorite stores online.
Mengram - AI memory API with 3 types: facts, events, and workflows
