Jena, 1806
On October 14, 1806, Napoleon destroyed the Prussian army twice in one day. At Jena and at Auerstedt, some fourteen miles apart, the heirs of Frederick the Great were beaten by a force that moved faster than their orders could travel. The Prussian soldier was the best drilled in Europe. He was also forbidden to think. Regiments stood in volley lines waiting for instructions from officers who were already dead or cut off. The rout that followed was worse than the battles. Fortresses with full garrisons surrendered to cavalry patrols because no commandant would act without an order from above, and above no longer existed.
The Prussians then drew a conclusion that most defeated armies never reach. The problem was not the soldier. The problem was a system that reserved all judgment for the top. Gerhard von Scharnhorst's Military Reorganization Commission rebuilt the army around a different premise, and out of that rebuilding came Führen mit Auftrag, leading by mission. The commander would state his intent and the outcome he needed. The subordinate would decide how. Two generations later Helmuth von Moltke compressed the entire philosophy into a single design rule. An order should contain only what the subordinate cannot determine for himself.
The reform took decades to become culture. Jörg Muth records that auftragstaktik was argued in Prussian military journals for thirty years before it hardened into standard practice. But when it hardened, it produced the most feared operational instrument in Europe. The decisive upgrade was never a weapon. It was a decision about where judgment should live, and Prussia's answer was as low as the technology of the age allowed. This paper extends that answer one step further than the Prussians could take it. The lowest unit of capability is no longer the regiment, the battalion, or the squad. It is the individual operator. The operator is the platform.
Capability moves down and doctrine follows
The pattern repeats across two centuries. A capability that lived at one echelon becomes cheap enough to live at a lower one. Doctrine, written around the old echelon, lags behind the hardware. Then it rewrites itself around the new lowest unit, and the armies that rewrite first win the opening battles of the next war.
The rifle did it to the musket line. Volley fire required a line because a musket could not be aimed, so the line was the unit of fire and the soldier was a component of it. When rifled barrels made individual aimed fire lethal at range, the line dissolved into skirmish order and the single soldier became a fire unit for the first time. The radio did it to the command net. When voice radio reached the squad in the Second World War, a corporal could call for fires that a colonel once had to arrange in person, and command nets flattened to match. Satellite navigation did it to the staff. By 1991 a private with a receiver could fix his position in a featureless desert better than a division staff of 1944 with every map in the inventory. Each device arrived as equipment. Each one ended as doctrine, because pushing capability down without pushing authority down produces soldiers who can see and shoot but may not decide.
The short history of this argument fits in one table.
| Capability | Old unit of employment | New unit | The doctrine that followed | |---|---|---|---| | Aimed fire | The line | The soldier | Skirmish and open order | | Voice radio | The battalion | The squad | Flattened command nets | | Precision navigation | The headquarters staff | The individual | Distributed maneuver | | Computing and AI | The command post | The operator | Not yet written |
American doctrine has already conceded the principle. ADP 6-0 defines mission command as empowering subordinate decisionmaking and decentralized execution appropriate to the situation. The words are Prussian in everything but language. What doctrine has not yet absorbed is the last row of the table. The most capable computer ever carried into combat did not arrive through a program of record. It arrived in the operator's pocket, bought at retail, and it arrived years before any institution decided what to do with it.
The proof came from Ukraine
In 2014 a volunteer group called Army SOS began building an artillery fire control application named Kropyva. No requirement document existed. No program office managed it. Soldiers used it because it worked. By the full-scale war, CSIS found that 90 to 95 percent of Ukrainian artillery units had adopted it as their primary fire control system. The application cut the interval from target acquisition to fire by as much as tenfold. A gunner with a tablet now did work that once belonged to a fire direction center.
Kropyva was not an outlier. Between 2014 and 2022 Ukrainian volunteer groups built eleven separate situational awareness systems and reached near-universal adoption among the forces they served. Dzvin, the official state-run command system, stalled in bureaucracy over the same years. Delta, the situational awareness platform that NATO observers later praised, began inside a volunteer drone unit in 2016 and became the de facto standard long before the Ministry of Defence formally took it over in 2023. The state did not field these systems to the force. The force fielded them to the state.
The pattern now runs through the whole Ukrainian military. The Army+ application serves more than 800,000 authorized military personnel and is replacing roughly 20 million paper requests a year, and the DOT-Chain marketplace lets brigades pick from more than 550 types of unmanned systems the way a unit once drew ammunition. Chatham House reached the structural conclusion. Ukraine's edge comes from decentralized innovation that lets frontline units adapt faster than any central authority could direct. And the same logic held under fire. When Russian electronic warfare degraded GPS-guided Western munitions, the fixes came from crews at the edge, iterating week over week against a live adversary.
American observers have started to say it plainly. A Modern War Institute essay on drone warfare argues that dominance will come from letting the squad experiment and fail rather than from perfecting a program in advance. That is auftragstaktik applied to materiel. Mission orders for hardware. The commander states the intent. The lowest unit figures out the how, and increasingly the lowest unit figures it out in software, on devices it already owns.
What the claim decides
The operator is the platform is a doctrinal claim, and doctrine is only useful when it changes decisions. So here is what the claim decides.
First, fielding is measured in individuals. A platform used to mean a hull. You counted ships and airframes and knew what you had fielded. When the platform is a person, the fielding question becomes how many operators woke up this morning with the capability on the device they carry. A brigade with 3,000 devices is 3,000 platforms, whether or not anyone has organized it that way. Most forces have not, which means most forces are carrying an unfielded fleet into the field every day.
Second, software is the capability increment. The March 2025 Secretary of Defense memo stated that software is at the core of every weapon and supporting system the department fields. Hulls change on decade timescales. Software changes on mission timescales. If the platform is the operator, the upgrade path is an install, and a force can be re-armed between one patrol and the next.
Third, adoption is the acquisition test. Kropyva never passed a milestone review. It passed the only test that has ever predicted battlefield value, which is whether soldiers in contact choose to use it. Usage under fire is operational test and evaluation run continuously and for free. An acquisition system that cannot read that signal will keep funding its own Dzvin while its own soldiers quietly standardize on something else.
Fourth, the team owns its intelligence. Data made by a team in contact is that team's tactical advantage before it is anyone's enterprise asset. When every observation must travel to a distant cloud to become useful, the team's own knowledge is held hostage by a link, and links fail exactly when the fight starts. The operator-platform keeps its intelligence where it was made, shares it deliberately, and destroys it deliberately when the mission ends.
Doctrine made software
EdgeLance exists to build the mission layer for the operator-platform. That is the company's whole position, stated without hedging, because a doctrine paper should say what its authors want. Someone will build the software layer that treats the operator as the platform. We intend it to be us.
The design follows the doctrine choice by choice. Intelligence runs locally, on hardware the operator already carries, because the operator-platform cannot depend on a link to a distant model any more than a rifle can depend on a link to a distant firing pin. Any capable model. Any capable hardware. The layer is indifferent to both, because the unit it serves is the operator and the operator does not get to choose the conditions of the fight.
Disconnection is the baseline condition, never the failure mode. Everything a team needs to plan, execute, and review a mission lives on the team's own devices and syncs when links happen to exist. The electronic warfare record over Ukraine settled the argument about which assumption survives contact. Security follows the same logic. The boundary of trust is the verified presence of the operator, never the health of a network, because a mesh heartbeat is exactly the kind of signal an adversary learns to kill first.
And mission data belongs to the mission. The team that generates intelligence owns it, works it into an after action review, and can burn it completely when the mission ends. Complete control of what a mission knows, held by the people who fought it. None of this is a feature list. Each choice is a doctrinal position rendered in software, which is why we describe the product as doctrine made software rather than software with a doctrine section in the brochure.
What the institutions do now
Acquisition already has its orders. The March 2025 memo made the Software Acquisition Pathway the preferred pathway and set Commercial Solutions Openings and Other Transactions as the default for buying under it. The memo's diagnosis was blunt. When software lags, "it is the Warfighter who pays the price." The remaining step is to treat adoption the way Ukraine treats it. Fund what the force already uses and sunset what operators abandon. Measure fielding in people. A program office that can tell you how many airframes it delivered but not how many operators opened the application yesterday is reporting the wrong number.
Training changes from operating equipment to exercising judgment. If every operator is a platform, then every operator is also a fielding decision and a very small acquisition office. Ukraine trains soldiers who modify their own tools at the front and feeds their answers back into procurement, with more than 370,000 survey responses already collected through the force's own application. A military that wants the same adaptability has to build it in peacetime and grade commanders on how much initiative their formations can absorb without breaking.
Command changes least of all, which is the point. Mission command doctrine already grants the authority. ADP 6-0 tells commanders to communicate intent and push execution down. What was missing was the means, because intent without local capability is only permission to be helpless. Local compute, local intelligence, and local data close that gap. The operator-platform is what mission command looks like when the technology finally catches up to the doctrine that has been waiting for it since 1806.
Prussia needed the destruction of its army at Jena before it would move judgment down the ranks. The evidence is cheaper this time. It sits in the published adoption numbers of a force at war, and in the pocket of every operator who already carries more computing than a corps headquarters took to Desert Storm. The platform has already been fielded. It walked into the formation on its own two feet.