Automating the Tedious: Unlocking Our Growth Strategy
In my role, I analyze data and guide strategies, without getting lost in the weeds of manual data entry. However, a significant part of my workflow involved just that. We had a crucial data source that gave us a starting point — a list of company domains — but it lacked the rich firmographic and technographic details required for meaningful analysis.
The Challenge: The ~400-Record Monthly Grind
The process was painfully manual and inefficient. For every single record, I had to search on the website for the company’s industry, funding status, location and recent news. This wasn’t a small task; we process around 400 records every month. The time spent on this repetitive cycle was immense, preventing me from doing the high-impact work I was hired for. I wasn’t analyzing data; I was just creating it.
This bottleneck was more than an annoyance; it was a strategic drag. I was looking for a way to automate the enrichment process and reclaim my time for analysis and action.
The Solution: An AI-Assisted Development Journey
I believe there’s a right way and a wrong way to use AI. The wrong way is to treat it like a vending machine: To open a chatbot and say ‘Make me a marketing plan’ and blindly go with the results.
The right way I’ve found is to think of Al as a sous-chef. It can prepare ingredients at high speed, handling the grunt work, but you’re the head chef who creates the recipe and tastes the final dish. Your job is to be skeptical, constantly refining what it provides. That back and forth is the part of the process that can tap into your creative process.
Step 1: Defining the Vision With Gemini
I started with a clear overview of what the application should do. Later, I worked with Google Gemini to expand on these ideas and define the more intricate details. This collaborative process was invaluable. It helped me think through complex scenarios I might have otherwise missed, like sophisticated error handling. For instance, what happens if an enrichment process is paused? How should it resume? Tackling these questions early on saved me from major headaches down the line.
Step 2: Building the Code With Cursor.ai
With a detailed project brief in hand, I turned to Cursor.ai, an AI-powered code editor. While I can do some coding, building a robust, production-level application from scratch was beyond my immediate skill set. Using the detailed outline I created with Gemini, I was able to use Cursor.ai to generate the foundational code for the project.
This process was a fantastic opportunity to hone my skills. I gained hands-on experience with a Node.js backend, a React frontend, connecting to a HubSpot MCP server, implementing batch processing for large files, and, most importantly, continuously improving the user experience by testing the app repeatedly.
How the CSV Enricher Works
The final product is a streamlined application that transforms a tedious manual task into a simple, automated workflow.
1. Upload & Parse: The user drags and drops a CSV file into the web interface. The back end immediately parses the file, deduplicates domains and prepares for enrichment.
Figure 1: Prompting the user to review the sample enrichment before updating all records.
2. Enrichment & Quality Check: The application takes the domain name from each record and uses a weighted search across various sources to find the company’s proper name, funding news, headquarters and other recent events. To ensure accuracy, it processes the first five records and then pauses, prompting the user to review the quality of the enriched data. This ‘human-in-the-loop’ step allows for tuning before committing to the entire file.
3. Finalize & Act: Once the enrichment is complete, the user has two choices:
-
- Download: Export the newly enriched data as a CSV file that can be used in other tools
- Sync to HubSpot: Automatically update company records in HubSpot via the MCP server integration, pushing the new data directly where it’s needed most.
Figure 2: Data has been enriched, and is ready for downloading or updating directly via the hubspot-mcp-server.
The Impact and The Wins: From Data to Strategy
What amazes me most about this project is not just the final tool but the empowerment that comes with it. By freeing myself from the repetitive, low-value task of manually updating spreadsheets, I can now focus on what truly matters: Taking action on the data.
This has led to significant, tangible wins for the business:
- Uncovering Actionable Answers: The application allows me to seamlessly work with mismatched datasets, finally connecting the dots. For the first time, we can visualize our user base by industry type. This was impossible when the industry field was just a blank column in a spreadsheet.
- Informing Key Strategy: We now have a clear, data-backed picture of which industries are leveraging our software. This insight is a game-changer, allowing us to build highly targeted industry-specific and account-based marketing (ABM) plans.
- Empowering Sales: This automated enrichment process directly empowers our sales team. They can now connect with companies that perfectly match our ideal customer profile (ICP), making their outreach more relevant and effective, which in turn drives real growth.
If you’re battling a similar bottleneck, start small, stay curious and don’t be afraid to experiment. If you’re a VictoriaMetrics user, check out our MCP server — it’s built to make enrichment as effortless as possible.