As more industries become dependent on data to drive decision-making, you may realize you need the capabilities of a decision intelligence platform. With this comes a common question: Build, or buy?  

Would you be better served by tasking your engineering and IT teams with the creation of a bespoke solution, that would ideally meet the needs of your company without any excess “frills”—knowing this may require lengthy lead-times, as well as an agile team of developers who can pivot as demands change? Or, would you be better to buy, gaining immediate access to a pre-built solution—but then be required to integrate workflows according to this framework? 

A ‘build or buy’ decision paradigm 

While every organization’s combination of needs, requirements, and capabilities is unique, we’ve seen that companies typically fall into three 'build or buy' categories. The category that your organization falls into is determined by various corporate factors such as industry, size, location, developer and data scientist population. Additionally it can be influenced by organizational requirements, such as whether your organization is regulatory facing, its data sources, risk appetite, legacy architectures, and its need to be on (or ahead of) the tech curve. Understanding these three categories will help you decide what option is right for your organization. 

1. Solution-first organizations should buy: These tend to be organizations that  just need a simple solution to operationalize their use case, and do not have the luxury (or burden) of armies of developers. These organizations can range in size, from the largest multinational mining and minerals organizations looking to optimize their supply chains to the smallest tax offices gathering evidence for tax fraud. They prioritize their product or service delivery over software IP. Time to value, simplicity, and low TCO are all more compelling priorities than complex software engineering. 

2. Platform-first organizations should us a mix of code and platform solutions: These include organizations such as leading banks, insurers, car makers, major defense research agencies, and telecom providers, which tend to develop, configure, and maintain modular pipelines with platform tooling, supplemented by bespoke code. They might also use frameworks sold or provided to them (e.g., as open source) by code-first organizations (see below), for example the open-source Kafka streaming platform which was developed at LinkedIn. 

3. Code-first organizations should build: In the finance sector, these organizations are typified by 'Formula 1' teams, systematic/quantitative hedge funds, and their highly computationally skilled bank peers, perhaps some leading security organizations, the FAANG organizations and the CSPs. For them, code solves problems, builds IP, and differentiates and builds brands, sometimes through open sourcing their code. 

A question of best fit 

To determine which category you best belong, you need to ask yourself a range of questions: In what industry would you most clearly place your products or services? `Is your organization a multinational corporation? What is your organization's perspectives around hiring in-house developers? Does your organization face stringent regulatory requirements, particularly around data—and are these geographically dependent? And of course, what are the budgets and timeframes for solutions you wish to deploy?  

These questions may even come down to the department in question:  

• In the financial services sector, a revenue-generating, highly competitive front office of a global bank or the pricing arm of a reinsurer might tend toward a code-first build approach. However, a regulatory-facing middle/compliance office in the same organization might be platform-oriented. And, a compliance office functioning in smaller regional and national retail banks would perhaps lean to a solution-first approach to get their AML and KYC work done. 

• In government, a “skunk works” division of a large defense contractor might tend to a mix of code and platform tooling, while a police force tends towards the solution that just identifies the criminal network. 

• In healthcare, drug discovery tends towards platform supported by code configuration, while a hospital delivering critical and outpatient services wants better scheduling—solution-first—in the event of a flu or virus outbreak. 

No two organizations are exactly alike, and their calculations around the benefits and drawbacks of building versus buying won’t be, either. 

Getting the right decision intelligence support 

For decision intelligence use cases—where software automates or augments strategic, tactical, and operational decisions—Quantexa can support you wherever you sit on the build/buy spectrum. 

For organizations that want solution-first approaches to get the job done, Quantexa provides: 

• End-to-end workflows from data to decision in AML, pKYC, Insurance, Risk Management and Customer Intelligence, for financial services, public service, insurance, telecoms and healthcare sectors 

• 99%+ entity accuracy for your AI and decisioning use cases 

• Accessible interfaces for data investigation, graph and network visualization, and operational scoring environments 

• Open interfaces to common applications such as spreadsheets, relational databases, BI tools and data stores, including support for necessary and novel data sources and third-party models, for example text data import (e.g. documents) and multiple languages 

Quantexa supports platform-first organizations through: 

• Unifying of data from across your data estate 

• Deploying platform capabilities appropriate to the modules across your workflows, including data unification, entity resolution and quality, graph generation and analytics, and scoring 

• Delivering and communicating entity resolved contextualized data where it’s needed 

• Providing knowledge graph generation and analytics tooling from within standard Python and Jupyter Notebook ecosystems 

• Operating on common, industry-standard platforms for maintainability and re-use 

• Providing APIs for integration across modular deployments 

• Offering multiple user interfaces for time-saving data, graph and score exploration, and visualization and analytics 

For code-first organizations, Quantexa provides developer-enabling tooling that: 

• Delivers and communicates true and contextualized entities across your pipelines 

• Provides efficient graph generation and analytics tooling for big data environments and Python ecosystems 

• Helps validate and benchmark your in-house models and scoring methodologies in conjunction with Data Science teams 

• Delivers best of breed Entity Resolution, graph and analytics tooling into your data engineering & management, data science and DevOps pipelines through APIs 

• Uses popular developer platforms for batch processing, search and streaming workflows 

Making the right choice for your organization 

Organizations may have a robust team of in-house developers—or, alternatively, may be prioritizing “lean” operations and rapid time-to-value—and their leaders may still have significant uncertainty about whether building, buying, or some combination is the right approach for their company.   

And this is understandable; within a rapidly evolving technological landscape and an ever more complex regulatory environment, CTOs and other company heads often feel they’re chasing a moving target when trying to settle on a Decision Intelligence solution that meets their needs. 

See how the right Decision Intelligence Platform can bring simplicity, productivity, and insight, yielding truly transformative outcomes. Get in touch with us to discuss your organization’s unique challenges, goals, and requirements, and to learn more about Quantexa’s Platform.