deep choice helps small and medium-sized companies to choose their best generative AI setup and to implement generative AI solutions into their companies.
For inquiries please send us an email to contact@deep-choice.ai
I. Generative AI Ideation:
We jointly explore what your company could do with (generative) AI and provide guidance in identifying and evaluating AI use cases. You will receive a report that reflects an evaluation of AI potentials for your specific company, including insights on what similar companies are pursuing.
II. Generative AI Tool Evaluation:
Once you know, what to do with AI, we jointly explore the how. We will introduce you to the admittedly chaotic landscape of AI solutions and services, discuss your preferences and guide you towards practical tools for your company’s situation (honest recommendations without bias to our tool ValueFlow).
For generative AI users:
ValueFlow:
ValueFlow is a generative AI platform for small and medium-sized companies that allows you to use generative AI in a compliant and controllable way – including a document store, document chat, AI workflows, company-specific templates or custom data collection.
For generative AI developers:
Abstract Web:
Abstract Web is a highly configurable web automation framework that lets users collect publicly available data via an intuitive CLI.
Features:
- Crawl websites
- Extract data from web pages
- Request APIs
- (Mass-)Download files
- Perform actions at webpages (e.g., click a button)
- State/cookie management (e.g., for logins)
- and more (custom tasks)
Notably, it is able to abstract away the complexity of these features by providing pre-defined models of web pages.
flowgram:
flowgram is a framework for “flowgramming” – building a progam as a flow, i.e. a sequence of steps. This has a lot of advantages; in particular that, flows are resumable at each step, stringifiable, and visualizable. Developers can write flows in TypeScript, with full IDE/IntelliSense support.
schemaSPOT:
schemaSPOT is a highly opinionated dev tool that allows to define resource object shapes at one SPOT (single point of truth) in an extended JSON Schema dialect and to automatically generate from them:
- A PostgreSQL database schema
- JSON Schemas and validators
- TypeScript typings
- …
This follows the DRY-principle (don’t repeat yourself) and separation of concerns, since specification is separated from implementation. schemaSPOT could also be seen as kind of ORM (object-relational mapper). Due to PostgreSQL support, schemaSPOT integrates well with Supabase.