Every connection below is one drop‑in step. No code, no API wrappers, no auth plumbing. Pick what you need; Daisy does the rest.
Pick the model that fits the job. Switch later without touching your workflow.
Plug in your API key and call any OpenAI model from any workflow node.
Best‑in‑class for long context and careful reasoning. Same interface as every other provider.
Vision, multimodal, and very competitive pricing. One config away.
All your AWS region + IAM rules apply. HIPAA‑eligible regions supported.
Drop in your Azure deployment name and region. Daisy handles the rest.
Run AI on your own GPU. No data ever leaves your network.
Build and demo without burning credits. Configure rules; Daisy returns deterministic answers.
Specialty embedding model great for technical and code‑heavy corpora.
Send a Slack message. Post in Teams. Page the on‑call. Always loops a human in the right channel.
Send formatted messages with buttons. Reply to a thread. Update an existing message.
Post in channels, send DMs, read message history. Auth via Azure AD.
Plain text, HTML, attachments, CC/BCC. Use any SMTP server.
Workflow fires on every new email. Parse it, extract attachments, kick off the rest.
Notify your own services with HMAC signing — recipients verify it's really from you.
Subscribe to a topic, fire a workflow on every message. Perfect for IoT and event streams.
The tools your team lives in. Daisy reads, writes, and reacts to changes.
Read, append, update, clear rows. Build dashboards or use a sheet as a tiny database.
Eight operations in one node: create issues, comment, transition status, JQL search.
Open issues, comment on PRs, file releases, react to webhooks.
Sync leads, log activities, push enrichments back to Salesforce records.
Create or update contacts, move deals, attach notes from your AI workflows.
Read database rows, create pages, append blocks. Build internal tools on top.
React to charges, sync subscriptions, automate dunning flows.
List inbox, fetch a thread, send replies. OAuth handled for you.
Find free slots, book meetings, react when an event is created or cancelled.
Your data, wherever it lives. SQL, files, object stores, vector indexes.
Parameterised queries. Connection pooling handled. Five operations cover everything.
Identical interface to Postgres. Swap the connection string, done.
Document operations with native filter / projection syntax.
Get and put objects, list a prefix, generate presigned URLs for uploads.
Same object‑store operations. Service‑account auth.
Containers and blobs through one interface, same as S3 / GCS.
Upload, fetch, list, share. Workflows that handle real documents.
Plug Qdrant in as your knowledge‑base backend when you outgrow pgvector.
Turn a spreadsheet into rows of data, or assemble one to ship to Sheets / S3.
Beyond text — every modality your AI workflows might reach for.
Convert audio recordings to text. Drop in a meeting; get the transcript.
Turn text into natural speech. Save to S3, attach to email, post to your app.
Create images from a prompt. Save to object storage or share inline.
Extract text from scanned PDFs and images. Steer the model with a prompt.
The little helpers you'd otherwise reinvent in every workflow.
Map, filter, join, reduce. The Swiss‑army knife for moving values around.
Space out external calls, throttle requests, give a system time to settle.
Pause until someone clicks Approve. Perfect for refunds, deploys, sensitive content.
No connector? Call the API directly with full control over headers and body.
Emit log lines and audit‑log entries from inside any workflow.
Call one workflow from another. Build pipelines out of reusable building blocks.
Daisy's plugin SDK is a tiny HTTP server interface. Wrap whatever API you need; declare a manifest; install. It shows up in the editor like any built‑in connector — auth, audit, secrets handled the same way.