Services > Data architecture

Data to inform decisions and power services

Making progress on major initiatives often comes down to one sticking point: ensuring data is accessible and usable. The security, privacy, quality, and shareability of data can be the determining factor in whether public services are successful.

Ad Hoc can help you find, organize, and use your data to inform decision making and streamline service delivery.

Selected data architecture capabilities

Data research and strategy

Map your system and plan for improvement

Data pipelines, management, and ETL

Get data in the right place and format for your needs

Data visualization and storytelling

Support decision-making and prioritization

Data infrastructure

Platforms for secure, efficient data operations

DataOps

Methods and tools to scale data operations

A people-centered approach to data

Your data’s only as valuable as it is useful. Ad Hoc applies its people-centered approach to help organizations learn what questions people are trying to answer with data, where data exists in the organization, and how to bring data together to inform decision making and improve services.

Ad Hoc’s data informed methodology is based on the industry-leading CRoss-Industry Standard Process for Data Mining (CRISP-DM) and is grounded in our firm belief that people are at the heart of any successful data-informed initiative.

OUR DATA-INFORMED APPROACH

20

billion

Medicare claims that support healthcare innovation through the Blue Button 2.0 API. See the story

CASE STUDY: ARTIFICIAL INTELLIGENCE

Innovative approaches to data challenges

Ad Hoc encourages our teams to bring creative solutions and leading technologies to our customer’s challenges.

Recently, one team used a natural language processing tool to bring the benefits of artificial intelligence to a data problem that had proven to be too complex for traditional methods.

The customer had a database of text entries that were approximately 30% duplicates. Because the entries all had very subtle variations, traditional deduplicating tools couldn’t identify them.

We used an open-source large language model to identify duplicates with an 84% accuracy rate.

We then connected the tool to the database input fields through an API to nudge future inputs to use standardized language and keep the data clean.

With this careful application of AI, we were able to unlock new insights for our customer and allow them to improve how they used data to make decisions.

More on applying AI to your data

Case study

VA COVID-19 Dashboard

With support from Ad Hoc, our prime contractor Oddball and our team partner Empirical Path, the VA quickly built a COVID-19 digital response dashboard to help measure the trajectory and success of their COVID-19 digital services. Learn more about our work.

Related approaches used for Medicare Plan Finder:

  • APIs
  • Data Pipeline
  • Data Visualization
  • DevOps

Case study

Platform Analytics for VA.gov

Ad Hoc, as a subcontractor to the Prime contractor, Oddball, created data products and provided training that help teams make more meaningful, data-informed decisions. Learn more about our work.

Related approaches used for Medicare Plan Finder:

  • Data Discovery
  • Dashboards
  • Training
  • Platforms
  • User Research
  • Analytics