Targeted Sales for Camera-Guided CNC Automation

Classic sales data is not enough for CNC automation. A record such as mechanical engineering, 50 to 200 employees, Baden-Württemberg says little about whether a company is a relevant SherpaLoader prospect.

The signals that matter are often not stored in databases. They are hidden on subpages, machine-park lists, job postings, news articles, and image descriptions. Which brochures and data sheets a company provides is another clue.

• Which turning or milling machines are used on site?
• Are specific machine models, controls, or machine types mentioned?
• Are there signs of small and medium-sized batches or frequent setup changes?
• Is the company hiring CNC specialists or investing in new machining centers?

Sales teams otherwise had to search for these signals manually, interpret them, and translate them into a relevant pitch. That took time, created inconsistent quality, and made it easy to miss technically strong prospects.

CERTAINCE built the AI sales assistant for this: a custom solution tailored to MAFU-SHERPA’s sales logic. It works with concrete buying signals for CNC automation instead of generic company enrichment.

The first step was to structure the relevant product knowledge: camera-guided part recognition, 3D camera and laser technology, short setup times, simple dialog controls, suitability for turning and milling machines, retrofit capability for existing CNC machines, and model sizes for different part weights and geometries.

• machine-park pages, part spectrum, and concrete CNC machine types
• small and medium-sized batches, frequent setup changes, and multi-shift production
• job postings for CNC operators, millers, or turners
• investment news, automation projects, and technical details on subpages and in machine-park lists

The solution reads company websites, relevant subpages, and search results. The findings are not only collected but interpreted directly: does the company fit technically, which machines or processes indicate demand, and which argument is strongest for this specific operation?

The AI sales assistant brings the whole workflow together. It researches companies and their machine parks from public web sources, scores the fit from 0 to 100, and backs that score with strengths, weaknesses, and concrete conversation openers. The same findings produce an email draft tailored to the role of the contact. The sales representative reviews the draft and sends it from Outlook with one click.

Behind it run around twenty individual AI steps, orchestrated centrally on Cloudflare Workers and tracked one by one, so the cost and the result of every step stay visible. A Chrome extension brings the research directly into the browser, and the Outlook integration puts the finished draft into the mailbox. The solution runs in production in MAFU-SHERPA sales today.

When the solution finds a machine-park page on a manufacturing company website, it evaluates the page for CNC-relevant terms. If a machining center is mentioned, the pitch can focus on milling automation, flexible loading, and additional spindle time.

Signals for CNC turning machines lead to stronger emphasis on automated loading and unloading, bar stock, and recurring turned parts. Small batches and varied parts shift the argument toward short setup times and camera-guided flexibility.

Open positions for CNC millers, CNC turners, or machine operators are recognized as a possible capacity constraint. Investments in new machining centers mark a timing signal where higher utilization and automation can be discussed together.

The research does not end with a summary. The solution translates the detected signals into concrete sales logic and adapts the argument to the company and the role of the contact.

For management, ROI, delivery capability, and investment security are central. For production leaders, machine availability, setup effort, usability, and process reliability matter more. For technical contacts, the outreach can go deeper into interfaces, part handling, and specific machine geometries.

• Your website mentions several CNC machining centers. In exactly these milling applications, flexible loading without rigid fixtures is worth a closer look.
• Your job postings for CNC operators show that shop-floor capacity is a topic. The SherpaLoader can relieve existing operators instead of scaling additional shifts only through more staff.
• Because you emphasize small and medium-sized batches, setup flexibility is likely a decisive factor. Camera-guided part recognition directly reduces this effort.

A self-contained application was built for a narrower task: finding the most likely LinkedIn company profile for each name in a company list. The app could process names entered directly or imported from CSV and Excel files, and the results could be exported as CSV or Excel.

That did not require an AI agent. A web app built on Next.js and Vercel evaluated Google search results supplied by SerpAPI. It did not open or scrape LinkedIn pages. The tool excluded subpages such as jobs and posts, normalized the public profile URLs, and selected the most likely match using a transparent score based on word matching and search-result position.

Uncertain matches remained reviewable: a debug view showed alternatives, partial scores, and the reason each filtered result was rejected. The component makes the architecture boundary visible. Where fixed rules can test what is correct, a fixed automation runs. AI enters only where machine parks, batch sizes, and investment signals need to be judged in context.

With the implemented solution, sales no longer starts from a long list of similar-looking companies. It starts from a prioritized list of technically relevant manufacturers.

Every lead answers three questions clearly: why does this company fit, which CNC information was found, and what should sales say next?

The largest effect comes from connecting research and execution. The solution finds CNC-relevant details, scores them against the ideal-customer logic, and uses the same findings immediately for outreach. The criteria of an experienced sales specialist become repeatable signals.