What artificial intelligence is
Artificial intelligence (AI) means programs that do things which until recently needed a person: they read a document, understand a question, recognise what is in a photo. They do not just follow fixed instructions. They learn from data, recognise patterns and make decisions.
You already use them. Siri or Alexa answer questions, send messages and make calls on a voice command, for example when you are driving and do not want to touch the phone.
Integrating artificial intelligence into CRM software
The CRM is the program where you keep your customers, your quotes and your conversations with them. When you build AI into the CRM, part of the routine work is done by the program: it reads, searches, summarises and suggests. The team decides.
AI does not only take over manual work, it also changes what the customer gets: answers at any hour through chatbots, offers and messages suited to each of them.
What AI does in the applications we deliver
This is what works in the CRMs and ERPs we have built:
- It reads supplier invoices. A scanned or PDF invoice is read automatically, header and lines, and prepared for goods receipt. The reading rules can be adjusted, and a person confirms before saving.
- It checks the goods receipt against the invoice. It compares what you ordered with what the e-Factura invoice says and flags missing or uncertain items.
- It reads estimates and contracts. An estimate from a PDF or a photo is split into labour and materials and matched to your catalogue. Contracts and products can be imported the same way.
- It answers questions about the company. You ask "which invoices has customer X not paid?" and get the answer with a link to the record.
- It searches for products by photo. You search by a photo or a description and find the item in the catalogue, as with a fabric distributor we work with.
- It recognises the product from a photo of the label. If the product does not exist, it creates it in the catalogue, supplier included.
- It writes text. Product titles, descriptions and labels for the website, and suggestions for social media posts.
In the CRM we use ourselves at Veziv, AI summarises meetings with customers and pulls out the tasks, prepares messages for leads based on the company's public data, suggests replies on WhatsApp and writes the first draft of the project specification.
All the features are described on the AI assistant in the ERP page. You can read about the CRM itself on the CRM software development page.
Types of AI features in CRM
In the CRMs on the market you will mostly come across:
- chatbots and virtual assistants, which answer customers at any hour;
- voice assistance, commands and searches given by voice;
- sentiment analysis, which notices from messages and calls whether a customer is unhappy;
- personalised recommendations, products suggested from the customer's history;
- lead scoring, so agents call the leads with the best chances first;
- predictions, who is ready to buy and who is about to leave.
All of them cut the team's repetitive work and make the relationship with the customer simpler.
The benefits of integrating artificial intelligence into a CRM
- Less repetitive work. Data is no longer copied from paper into the program.
- Messages and offers suited to each customer. Based on their history, not at random.
- Faster answers. For customers, and for you when you ask something about the company.
- Decisions based on data. Machine learning algorithms can anticipate customer needs, so you react in time.
Challenges and solutions
Integrating AI into a CRM also has real problems. This is how we deal with them:
- Data protection. Access to data goes by roles and permissions, the same as in the rest of the application. See Security and audit.
- Wrong results. Nothing the AI reads goes into the system without a person checking it. The assistant's answers come with a link to the source, so you can check them.
- Instructions hidden in documents. The text in invoices or contracts is treated as data, not as instructions for the AI.
- Costs. AI usage can be logged for every request or scan, so you know what it costs you.
- The link to what you already have. The AI works on top of the data in your CRM and ERP, not in a separate program.
Artificial intelligence programs
The simplest example is a chatbot that talks to users following set rules or what it has learned from earlier conversations. Other examples: facial recognition, which identifies people in images, and machine translation, which translates text between languages on the spot. We use machine translation in the Pigmento app, where content translates itself when you save it.
Want to see what AI can do with your company's data? Book a demo or see our projects.