David ChavezFebruary 27, 2020

Building vs. Buying: Which AI Approach is Best?

  • Home
  • AI
  • Building vs. Buying: Which AI...

Over 80% of executives believe failing to get on board with artificial intelligence (AI) now will cost their organization for the next decade, according to Vanson Bourne. AI is driving revolutionary capabilities across the enterprise with departments like sales, marketing, HR and finance leveraging the technology for some of their most essential functions. As a result, organizations have increased their investment in AI by 23% over the last 12 months. But how are they investing?  

There are two paths to AI investment: building solutions in-house or buying from a third-party vendor. Vanson Bourne’s data shows that 24% of organizations’ current or planned AI solution is or will be exclusively developed in-house. The same amount (24%) plan to have their AI solution fully developed by a third-party. The majority (52%) are looking to a mix of in-house and third-party AI development.

What are the pros and cons of each investment option, and what does a blended approach look like? Let’s get into it…


  • PRO: Complete control over AI solution
  • CON: Development may be affected by skills shortages (74% of organizations struggle with a lack of understanding about AI and/or a lack of skills in-house to facilitate adoption)
  • PRO: Better control over budget (insufficient budget is a key barrier to adoption for 31% of organizations)
  • CON: Sole responsibility for AI rollout including ongoing maintenance and management; must have dedicated personnel in-house
  • PRO: Work at your own pace
  • CON: Rollout of AI may be slower, which can stall modernization/improvements
  • PRO: Better more accurate training data and bias reduction
  • CON: More responsibility to document how your AI is working and effective in an environment where lots of regulation will change


  • PRO: Get high-quality support through the entire lifecycle of your AI rollout (a top priority for 57% of organizations soliciting support from a third-party)
  • CON: You’ll never have complete control over your AI rollout (could be important for mission-critical features/functionalities)
  • PRO: Faster rollout of AI adoption (another top priority for organizations soliciting the help of a third-party)
  • CON: Potential difficulty finding vendors that can handle specifics that are best understood internally
  • PRO: Easily assess what parts of your strategy are working or not (the ability of a trusted expert to measure the effectiveness of your AI rollout)
  • CON: You may not have enough control over your data access, and derived data is visible and potentially owned by the 3rd party
  • PRO: Due to breadth of experience and implementations, the 3rd party should be able to address evolving security threats to AI systems
  • CON: Maybe they aren’t so good at the previous or didn’t prioritize it highly enough


Organizations want control over their AI rollout, which is gained by in-house development, but research suggests a skills shortage that necessitates the need for third-party support. This is where we see the benefits of an application ecosystem approach. In this way, organizations can maintain control over their AI rollout while leveraging powerful solutions from global market leaders like Google, Salesforce, Verint and (cough, cough) Avaya.

An application ecosystem environment provides the openness organizations need to seamlessly integrate third-party apps and features—both off-the-shelf, ready-to-use AI solutions and backend technologies—for developing their own unique AI solutions (i.e. bot, smart routing, conversational intelligence, biometrics).

In any event, the adopting company will have to evaluate a legal and regulatory landscape that will vary by geography and country. This environment will evolve over time, sometimes quickly and potentially unexpectedly. Some AI projects will not deliver what was promised, and will need re-evaluation. Other projects may be subverted by forces outside of the control of the business and even the 3rd party, security matters come to mind. So even as many businesses approach this area choosing to mix inhouse with third party, they may find that the combined drawbacks are significant enough to slow desired progress down.

Internally collected data can be used in conjunction with these third-party apps and features to improve customer and business outcomes, from connecting with customers to monitoring IT infrastructure to tracking financial analytics to performance management for HR.

The best benefits of an application ecosystem are seen within the contact center, an area of business that 94% of organizations believe can be transformed by effective AI. Consider Avaya’s application marketplaces—Avaya DevConnect and Avaya A.I.Connect—that include integration partners like Google, Nuance, Sestek, Salesforce and more.

The general availability of Avaya integration with Google Cloud, for example, allows organizations to deploy powerful, simplified, AI-enhanced communication and collaboration solutions that transform the customer and agent experience. These can include virtual agents for reducing handle time and increasing accuracy or conversational topic modeling for leveraging real-time visibility of topics. Integration with Nuance provides advanced speech enhancements for organizations to better personalize customer interactions and create more frictionless service experiences. The possibilities are truly endless.

Even with all the money in the world, 85% of organizations agree that they will need third-party support as AI becomes more prominent. Gartner predicts that by 2025, the average contact center organization will be exploiting the benefits of an application ecosystem to better equip staff and enhance service.

Explore our DevConnect and A.I.Connect marketplaces to see what AI solutions you can start implementing within your contact center (as in right now, today).

Building vs. Buying: Which AI Approach is Best?

David Chavez

David Chavez is a Vice President in the Office of the CTO, responsible for Architecture and Consulting. Having 25 years of experience and holding 76 US patents, David is responsible for Avaya's award-winning and market-leading IP communications architecture and the creator of the Avaya Aura(R) architecture. He has a B.S. in Computer Science and Mathematics from New Mexico State University and an M.S. in Mathematics from Colorado State University, with some executive education from Stanford Graduate School of Business.

Read Articles by David Chavez