Services
Custom software, AI systems and hosted products, for research centers, government agencies and companies. To see finished projects, go to our work.
What we build
Every project ends with the system in use and your own staff able to run it.
- Custom software
- A system you own. We build it on your own data, document it, and train your staff to run it. Examples: one data warehouse for a dozen city systems, and a travel insurance sales platform in eight countries. Custom software
- AI systems
- Software that does the repetitive part of skilled work: data entry, follow-up and reformatting. It shows its sources and asks a person before any decision that matters. We built a proposal advisor for a National Science Foundation program, and we run our own company on the same approach. AI systems
- Products
- Two systems we built for clients and now license by the year: SimplyScholar for research centers, and service alerts for city public works departments. We host, maintain and support them. Products
What we are good at
We have solved these six problems for clients in government, research and business. Each one links to a project.
- Connecting systems that were not designed to work together
- We connected a dozen DC Department of Public Works systems, none of them built to share data, through one integration layer. Every connection to an outside system goes through that layer, so when a vendor changes, only one connection changes. DC Public Works, since 2016
- Fast reporting from one data warehouse
- One data warehouse for the whole operation. At DC Public Works, reports that arrived weekly as a spreadsheet now update within minutes. How the warehouse works
- AI answers that show their sources
- The EPIIC Assistant, our proposal advisor for a National Science Foundation program, gives stored expert answers to common questions and AI answers based on the program guidance for the rest. We also built an instrument finder for the Materials Innovation Platforms that links every claim to its source page. People act on an answer only when they can check it. Software and AI for research centers
- Experience with sensitive data
- We built and ran Health Insurance Portability and Accountability Act (HIPAA) infrastructure for the Maryland Department of Health. It met the Center for Internet Security (CIS) Level 1 benchmark and passed a state security review. We also built student case management for DC Public Schools, covered by the Family Educational Rights and Privacy Act (FERPA). COVID-19 response systems
- Large message volumes
- 45 million messages sent and received for two state health departments. About 100 million readings from a 200-vehicle city fleet. Service alerts for DC residents, who know them as MyDPW. Service alerts
- Performance tuning and load testing
- We moved maryland.gov from Windows to Linux, secured it, then load tested and tuned it to handle 150,000 requests a minute. It took four working days, and it was ready before the governor's press conference. How we rebuilt it in four days
By type of organization
How you buy, and who approves, differs by sector. Each page covers yours.
-
Companies
AI and software for business operations
Systems that take repetitive work off your staff, answers from your own documents with sources, and help to finish an AI project that stopped after the demo.
For companies
-
Research centers
Less administrative work for research centers
For National Science Foundation and Department of Energy research centers: SimplyScholar for center records and websites, instrument finders for a whole program, and proposal advisors.
For research centers
-
Government
Software and AI on contracts you can use now
Resident request intake, document work, and connections between systems bought separately over many years. Available through Maryland's Consulting and Technical Services (CATS+) contract and SAM.gov.
For government
Two fixed-price ways to start
Each review costs $12,000, takes two weeks, and ends with a written report. Neither one commits you to further work, and neither fee is credited against a build. If neither fits, book a 30-minute call and we will find the right first step together.
- AI Workflow Review
- For manual work and no plan yet. Bring two or three tasks your staff do by hand. For each one, we tell you what it would take to build, what it would save, what could block it, and which to do first. Up to three tasks, and about six hours of your time.
- Production Readiness Review
- For an AI project that is stuck after the demo. We read the code, talk to the people who must approve it, and give you a written recommendation: launch it, fix it, or stop. One project or system, and about four hours of your time.
Not sure which of these you need?
In a 30-minute call, we will tell you if it is worth building, what it would cost, and how long it would take.