The Heilmeier Catechism for AcqBot

As stated on the DARPA website "DARPA operates on the principle that generating big rewards requires taking big risks. But how does the Agency determine what risks are worth taking?

George H. Heilmeier, a former DARPA director (1975-1977), crafted a set of questions known as the "Heilmeier Catechism" to help Agency officials think through and evaluate proposed research programs."

The genus of AcqBot started at DARPA, so it's fitting to put AcqBot in the context of "the Heilmeier Questions" that DARPA Program Managers still have to answer to gain approval for their new groundbreaking programs.

1. What are you trying to do? Articulate your objectives using absolutely no jargon.

There are few, if any use cases in the federal government better suited for the application of AI than acquisition. The process is structured, process-driven, involves an immense amount of paperwork and data, and there is a mountain of legacy artifacts. We are seeking to make that AI application turnkey.

2. How is it done today, and what are the limits of current practice?

Today the process of acquisition is highly manual. In spite of the middle-tier acquisition framework that removed or streamlined some of the steps, the steps still basically involve writing things on paper, filling out forms, and volleying them between government personnel and industry counterparts.

3. What is new in your approach and why do you think it will be successful?

We built a streamlined platform that builds automation around current workflows, mirroring current processes so that users don't have to change their processes or learn new ones, they do the same things, just faster. We also built generative AI into the back end of each process and step, so users don't have to learn cumbersome prompt engineering, utilizing AI becomes as simple as filling in some fields, and pressing a few buttons.

4. Who cares? If you are successful, what difference will it make?

In the end every segment of the government will care, because nothing happens without contracting, from ballpoint pens to 6th gen fighters. But the first people who will care are the beleaguered acquisition workforce who are constantly being asked to do more acquisition actions with fewer hands.

5. What are the risks?

Workforce rejection of automation and AI in favor or trusted, manual, legacy approaches.

6. How much will it cost?

To-date stakeholders in OSD, Air Force, Army, and Navy have invested in non-reoccurring engineering to deploy the present platform. further organization tailoring and deployment ranges from free simplified acquisition threshold purchases.

7. How long will it take?

Deployment can take as little as a week.

8. What are the mid-term and final “exams” to check for success?

1. MVP deployment - DONE

2. Operational deployment - DONE

3. IL5 ATO - DONE

4. .mil DNS - in process

5. IL6 ATO

6. Service-wide enterprise deployment

7. IL7 ATO