FlowDoo · Connect & Automate
Status: AvailableConnect. Transform. Automate.
More than 280 visual steps for structured and unstructured data — from databases and spreadsheets to PDFs, free text and audio — with your files and connections kept in one secure workspace.
- What it is
- FlowDoo is a visual pipeline builder. You place steps on a canvas, connect them, preview the result at any point and run the flow on demand, on a schedule or when another system calls it. Its 280+ steps cover structured data (databases, spreadsheets, files) and unstructured data (documents, PDFs, free text and audio), and everything you build, upload and connect lives in your own workspace.
- Who it is for
- Analysts, operations teams and data engineers who move and clean data between systems and are tired of doing it by hand or by script.
- Status
- Available. Generally available.
The problem
Most data work is still exports, spreadsheets and scripts nobody else can read.
The same file is exported, cleaned and re-keyed every week because automating it meant a development project.
Integration tools move data but stop short of documents, PDFs and the files people actually work in.
A script that works is a script only its author understands; when it fails at night nobody knows why.
Core capabilities
What FlowDoo does.
Capabilities marked Planned are on the way and not yet in the product.
Ingestion
Read files and whole folders (CSV, Excel, JSON, XML, Word, PDF) and nine databases — MySQL, SQL Server, PostgreSQL, Azure SQL, Oracle, MongoDB, Snowflake, BigQuery and DuckDB — plus Google Sheets, Google Drive, Salesforce, PowerSchool, S3, FTP/SFTP, SharePoint, email inboxes and REST APIs.
Transformation
Reshape data with pivot and unpivot, group by, formulas, type conversion, ranking, surrogate keys and slowly changing dimensions.
Joins and lookups
Join, look up, union, intersect, subtract and merge tables — by rows or by columns.
Data quality
Profile data, validate schemas, quarantine rows that break your rules, find outliers and near-duplicates, standardise values and verify addresses.
Parsing
Flatten JSON, parse XML, HTML, logs, vCards and calendar files, and pull values out with regular expressions.
Unstructured text
More than 50 text tools: clean, count, sort and de-duplicate lines, extract emails, URLs and numbers, encode and decode, and format JSON — on whole documents, not just cells.
Audio analysis
Transcribe speech with Whisper, tell who spoke when, detect voice activity, classify sound events and extract features such as MFCCs and spectrograms.
Documents and PDFs
Merge data into Word templates, then generate, merge, split, watermark, compress, encrypt, fill and OCR PDFs, or extract their tables, text and images.
Reports
Lay out paginated reports with headers, tables, fields, images, barcodes and QR codes, or produce an HTML report.
Analytics in the flow
Summarise, cross-tabulate and profile; cluster, run regressions and decision trees, forecast time series, raise anomaly alerts, score sentiment and extract entities.
Security steps
Hash, encrypt and decrypt, mask, redact and tokenise sensitive columns, and neutralise spreadsheet formula injection before files leave the flow.
Write back
Append, replace or upsert into any of the nine databases inside a transaction, or send results to Sheets, Salesforce, VizzDoo, cloud storage, email or an API.
Automation and triggers
Run on a schedule, on a signed webhook or when a file arrives; loop over files and rows, reuse sub-flows, retry, handle errors and pause for an approval.
Notifications
Alert the right people by email, Slack or Microsoft Teams when rows are quarantined or a run needs attention.
Personal workspace
Your own storage, shared by every Doo product: upload files and folders, organise them and drag any of them onto the canvas to start a flow.
Saved connections
Save a database or platform connection once and it is ready in every flow. Share it with a colleague and they can use it without ever seeing the password.
Encrypted flow files
Exported flows (.fdt) are encrypted and tamper-evident, so mappings and connection details cannot be read outside DooHorizon. Import them back with one click.
Desktop agent
Let scheduled flows read and write folders on your own computer and run approved programs there — over an outbound connection only.
- Planned
Model execution
Score rows with a model deployed from MLDoo as an ordinary pipeline step.
- Planned
AI execution
Hand a step to an AIDoo task or agent and continue with its answer.
How it works
From first connection to a running result.
Step 1: Connect your sources
Pick a file from your workspace, or a database, spreadsheet, CRM or API connection you saved once. Credentials are encrypted on the server and never returned to the browser.
Step 2: Build the flow on a canvas
Add steps to match, clean, join and validate. Each step shows a preview computed by the same engine that will run the pipeline.
Step 3: Decide what happens to the result
Write to a database, a spreadsheet, a document, a file — or straight into a VizzDoo dataset.
Step 4: Let it run
Run on demand, on a schedule, on a webhook or as a chain of flows. Every run keeps a log of rows in, rows out and what it could not read.
Example workflow
What it looks like in practice.
A typical pipeline: read two sources, standardise and join them, check the rows that do not fit, then write the clean result where the business needs it. The Model step is planned.
- Input
- Transform
- Join
- Validate
- Model
- Decision
- Output
Part of DooHorizon
How FlowDoo connects to the platform.
The pipeline you draw is the pipeline that runs — every step previews against real data before anything is written.
VizzDoo
A FlowDoo pipeline can write its result directly into a VizzDoo dataset, so a dashboard is fed by the flow that cleaned its data.
About VizzDooMLDoo
Prepared data becomes a versioned MLDoo dataset; scoring rows with a deployed model from inside a flow is planned.
About MLDooAIDoo
Status: Coming SoonAIDoo is designed to call FlowDoo flows as tools, so an agent can take a real, logged action.
About AIDooGovernanceDoo
Status: Coming SoonPipelines are designed to register with GovernanceDoo as governed assets with an owner, lineage and run history.
About GovernanceDoo
Security and governance
Built to be trusted with real data.
Designed with enterprise security and governance requirements in mind. No certification is claimed that has not been obtained.
How DooHorizon approaches security- Every page and API route requires a signed session; server paths are confined to your workspace.
- Connection credentials are encrypted with AES-256-GCM and are never sent back to the browser; a shared connection never reveals its secrets.
- Exported flow files are encrypted and authenticated, so only a DooHorizon server can open them and any tampering is refused.
- Webhook triggers only start a flow for callers that sign the request with the flow's secret.
- Outgoing connections are checked against private-network ranges, and databases are read-only unless an administrator allows writes.
- Your files live in your own workspace, shared only with the licensed users you choose; access is checked on every read, so revoking it takes effect at once.
Use cases
What you can build with FlowDoo.
Automate recurring reporting
Replace the weekly export-clean-paste routine with one flow that runs before the meeting does.
Sync business systems
Move records between a CRM, a student or ERP system, a spreadsheet and a database on a schedule.
Process documents at volume
Extract tables from PDFs, merge data into templates and produce finished documents without opening them.
Prepare data for analytics and ML
Standardise and validate data once, then feed dashboards and models from the same pipeline.
Turn recordings into data
Transcribe calls and meetings, split them by speaker and land the text in a table you can search and analyse.
Clean up unstructured text
Strip, normalise and mine pasted or exported text — emails, logs, web pages — before it reaches a database.
The product
A look at FlowDoo.
- InputOrders.csv · CRM
- TransformClean · convert types
- Joinon customer_id
- Validate3 rules
- ModelPlanned step
- DecisionRoute by result
- OutputDatabase · VizzDoo
| customer_id | name | order_date | check |
|---|---|---|---|
| C-1042 | Northwind Ltd | 2026-09-30 | Valid |
| C-1043 | Aster & Co | 2026-09-30 | Valid |
| C-1044 | — | 2026-09-31 | Needs review |
Prefer the real thing? Open FlowDoo — it is free to start.
FAQ
FlowDoo: common questions.
Do I need to write code to use FlowDoo?
No. Pipelines are built visually. A SQL step and formulas are available when you want them, but nothing requires them.
Where does my data go when a flow runs?
Flows run on the FlowDoo server in a private workspace. Files are deleted automatically after the retention period, and the platform can be self-hosted when data must stay on your own infrastructure.
Can a flow use files on my own computer?
Yes. A small desktop agent lets scheduled flows read and write folders you choose, without opening an inbound port.
Where are my files and connections kept?
In your workspace: your own storage with folders, uploads and saved connections, shared by FlowDoo, VizzDoo and MLDoo. You decide who else can see each item, and whether they can only view it or also edit it.
Can someone read a flow I export?
No. A .fdt file is encrypted on export, so the steps, mappings and any connection details inside it cannot be read in a text editor. It opens only by importing it into DooHorizon.
Can I use FlowDoo without the other Doo products?
Yes. FlowDoo works on its own; the connections to VizzDoo and MLDoo are there when you want them.
Start with FlowDoo. Expand when you are ready.
Start with one use case. Expand across the ecosystem.