Formula Bot scales data connectors and enterprise workflows as a solo founder with n8n

Background

Formula Bot is an AI powered data analytics platform founded and run by David Bressler. The product allows users to upload or connect data from a wide range of sources, from static files like PDFs and Excel to live systems such as Google Analytics and databases including BigQuery. All analysis happens through natural language.

Users can ask questions in plain English and automatically generate insights, charts, spreadsheets, and even data science models. More recently, Formula Bot expanded into automated reporting and auditing, enabling customers to schedule recurring reports delivered daily, weekly, or monthly.

David operates as a solo full time founder, supported by two developers. He has a strong background in data analytics, databases, and schemas, but does not consider himself a traditional software engineer. To build and scale Formula Bot, he relies heavily on low code and no code tools, using Bubble for the front end and database layer, and n8n as the core backend orchestration engine.

As David explains, “I’m not a coder. I’m logical. My background is data analytics, and n8n just makes more sense to me than building everything in code.”

Challenge

As Formula Bot evolved, so did the complexity behind the scenes. David needed to:

• Integrate many external data sources, including databases, analytics platforms, and cloud storage
• Orchestrate complex data transformations and AI workflows
• Run long lived workflows without timeouts
• Securely manage user data, particularly within AWS

Initially, much of this logic ran on AWS Lambda and custom code. For a founder who was not a developer by training, this introduced serious friction. Workflows were opaque and fragile. David could often make things work, but he lacked confidence in their correctness or durability.

"Just because something works doesn’t mean it’s right,” he explains. “I would build something, and then something else would indirectly break later."

At the same time, Formula Bot struggled to scale its data connectors. Each new integration was effectively a one off project, using APIs and function calling through tools like OpenAI’s Code Interpreter and agents. These flows were difficult to reuse, slow to build, and impossible to template. As a result, Formula Bot stayed limited to only a few connectors for nearly a year.

Bubble also introduced a hard constraint. Its API connector enforces strict execution time limits. With AI reasoning models and heavy data processing, many workflows exceeded those limits.

"There is a limit with Bubble in terms of how long an API can run,” says David. “It’s not long enough for a lot of the functionality on my site."

This became a blocker for high value use cases such as:

• Web scraping large pages and tables
• Converting complex PDFs into structured CSV or Excel files
• Running multi step AI reasoning workflows

As David pushed Formula Bot toward enterprise customers, the problem intensified. Enterprise users required deep integrations with systems like BigQuery, Snowflake, Microsoft SQL, and Google Analytics. Without a flexible backend orchestration layer, Formula Bot risked being constrained by integration work and technical complexity.

Solution

To address these challenges, David adopted n8n as the core backend engine for Formula Bot. Today, roughly 60 percent of Formula Bot’s workflows run in n8n.

At the heart of the architecture is a central orchestration agent built in n8n. This agent receives user requests and dynamically routes them to the correct connector workflow.

"I essentially have one orchestration agent that receives the request,” David explains. “Based on that, it decides which tool or agent to call."

The flow works as follows:

1. A user uploads or connects a data source through the Bubble front end.
2. n8n retrieves credentials and metadata by calling Bubble’s API, including schema information stored securely in Bubble.
3. An AI agent generates and executes the appropriate SQL or API queries based on the schema and user intent.
4. Results are processed, transformed, and stored in AWS S3, then returned to the user as structured files.

Once this pattern was built, it became highly reusable. Around 90 percent of the workflow is templated, regardless of the underlying system.

“Now that the flow is solid, adding new connectors is much easier,” says David. “We swap the system we talk to, whether it’s BigQuery, Snowflake, or Google Analytics, but everything else stays the same.”

n8n also solved Bubble’s API timeout limitations. Instead of running long operations inside Bubble, Formula Bot now uses n8n webhooks as an asynchronous backend. Bubble receives an immediate response, while n8n executes the heavy processing in the background.

This pattern powers features such as  web scraping agents that construct structured requests for third party scraping APIs or PDF to Excel conversion, where n8n extracts unstructured content, converts it to JSON and CSV, stores it in AWS, and returns the final file.

Results

Moving backend orchestration into n8n delivered significant gains across development speed, product capability, and business performance.

Connector development accelerated dramatically. Before n8n, adding a new data connector took about a week. With reusable n8n workflows, the process now takes around a day and a half, mostly for validation and testing.

This enabled Formula Bot to expand rapidly into enterprise grade integrations, directly supporting its move upmarket.

“Having way more connectors is broadening our business opportunities,” David notes.

The time savings for a solo founder have been substantial. David estimates 20 to 30 hours saved per month, and hundreds of hours overall, by building workflows in n8n instead of maintaining custom Lambda functions and ad hoc API code.

n8n also removed the ceiling imposed by Bubble. Formula Bot can now:

• Run workflows lasting up to 10 minutes
• Chain multiple APIs and reasoning models
• Deliver advanced features like enterprise data scraping and multi database querying

This level of capability would typically require a much larger engineering team.

Ultimately, the impact shows up in the business itself. “It’s helped me build a stronger product,” says David. “Growth, profitability, and product feedback are all up. Everything is running at full speed.”

For a solo founder supported by two developers, n8n has been the key enabler in turning Formula Bot from a novelty Excel tool into a robust AI powered data analytics platform.