Custom build methodology

Custom build

How AIM sizes a system that is not in the catalog, and why the hour range, the TCO dollar, and the staffing dollar stay separate.

Freedom AIM eagle and gear mark

What custom build is

Catalog recommendations answer which existing product fits this organization. Custom build answers the question that comes first when no such product exists. The organization describes the system to be created — capabilities, integrations, non-functional requirements, and how novel the work is — and AIM produces a cited hour range for planning.

The same inputs produce the same range. The coefficients and the sources they cite are stored with the estimate, including the phase split and the labor mix used for that run. A later change to AIM's current defaults does not rewrite an estimate that was already saved.

Coherence check

After the scope is saved, AIM records contradictions already present in that scope: real-time throughput without a real-time module, restricted data without audit logging, a buy signal that has not been resolved. When recommendations are generated, one review lists assumptions, unknowns, and objections, and it may propose a module change. A person accepts or rejects that proposal. The review does not narrow the hour band and does not change a RAO score.

What the recommendations mean

On a custom-build assessment, recommended products are the enabling stack: the platforms, data services, and integration tools the organization would build on. They are not a substitute for the system being created. A buy check can name catalog products worth a second look. That check is advisory. The cited build estimate remains the quantity.

Hours, then two different dollars

The hour band is the quantity. It is a planning range with a low, a mid, and a high. It is not a bid, and it is not anchored to awarded-contract prices.

Total cost of ownership converts the mid hours with a single planning rate of $150 per hour, and carries 15 percent of that implementation labor as sustainment from year two onward. Staffing converts the same hours into full-time equivalents at 140 productive hours per FTE-month across the stated calendar duration, then prices those FTEs with the sector labor model used everywhere else in AIM — market rates, staff augmentation, or contractor rates, including GS and SCA where the sector calls for them. A federal IGCE uses its own GSA and BLS rate path. These dollar figures describe the same hours through different rate books. They will not match.

Benchmarks on the same assessment

A USASpending benchmark can still appear. It compares the assessment's modernization fingerprint with awarded programs and is useful as program-scale context. That comparison is not a calibration of the cited hour pool. AIM discloses that there is no awarded-contract set of greenfield custom builds behind the hour range.

Firmware is a separate model

Firmware and FPGA work is sized on its own, with an assurance level in the model. It is not folded into the software hour pool. The result is advisory. Named humans and licensed contractors perform the build.

After the decision, in PULSE

When a PULSE is created from a custom-build assessment, AIM links the estimate that was current at that moment. When every epic is attested done, AIM records that planned hour band next to what the epics actually did: how many finished on time, and how many carried a blocker. AIM does not invent a count of hours the team worked. On-time delivery and blockers do not show whether the hour estimate was accurate, and they do not change the sizing coefficients. An organization can later record the hours it actually spent. That record is kept for review. It does not retune the next estimate.

Custom Build | Freedom AIM