AI agents

Ten agents, a twin and an orchestrator. Each one bounded.

Every Powdrix agent owns a stage of the additive floor, a declared tool set, the signals it watches and the parameters it is allowed to move inside the qualified envelope. Five of them ship as named products. The orchestrator plans, arbitrates and routes approvals; none of them acts outside the policy engine, and none of them can leave the envelope without a human.

10 agents · 1 twin · 1 orchestratorBounded tool setsEvery action logged

run_7c2e4b · agent graph · PBF-2 · scenario 10 nodes · 1 approval gate
01 · orchestrator ingest.job ✓ SUCCEEDED 02 · parttwin twin.simulate ✓ SUCCEEDED 03 · prep_parameter prep.release ✓ SUCCEEDED 04 · recoat_powder recoat.check ✓ SUCCEEDED 05 · melt_layer melt.stabilise ✓ SUCCEEDED 06 · thermal_distortion thermal.compensate ✓ SUCCEEDED 07 · orchestrator approve.human ◆ APPROVAL 08 · robot_handling handle.depowder ✓ SUCCEEDED 09 · defect_inspect inspect.ct ✓ SUCCEEDED 10 · quality_conf release.qualify ✓ SUCCEEDED

Illustrative scenario: PBF-2 · Ti-6Al-4V bracket family, 24-part build, 23 of 24 parts released, one held

Agent roster

Ten agents, a twin and an orchestrator

Each agent owns a stage of the additive floor, a bounded tool set and a metric the floor already measures. They negotiate for shared machines and cells through the orchestrator, never directly. Five of them ship as named products.

agent.prep_parameter

Prep-and-Parameter

Build preparation, orientation, supports, nesting, process parameters

Proposes orientation, support, nesting and process-parameter recipes (laser power, speed, hatch, layer) for a build, scores them against density, geometry and takt targets with the part twin, and releases the winning recipe to the printer only inside the OEM-approved parameter envelope and after sign-off.

slicer.read_buildtwin.rank_recipesprinter.release_recipe

agent.melt_layer · shipped as Meltrix

Melt-and-Layer

In-situ melt-pool, layer and energy control

Fuses coaxial and off-axis melt-pool video, layer and recoater images, photodiode and thermal streams at high frame rate, classifies and predicts porosity, lack of fusion, keyholing, spatter and recoat faults seconds ahead, and closes into bounded, operator-graduated laser and energy control to hold the pool stable.

holoscan.fuseprinter.trim_powergenealogy.tag_layer

agent.recoat_powder

Recoat-and-Powder

Recoating, powder-bed uniformity, powder quality and reuse

Watches recoater images and bed uniformity every layer, detects streaks, short feeds and spreadability faults before the next layer is fused, tracks powder lot, reuse count and quality, and recommends or applies bounded recoat and dosing corrections.

recoater.read_imageprinter.adjust_dosepowder.lot_record

agent.thermal_distortion

Thermal-and-Distortion

Thermal fields, residual stress, distortion

Tracks the thermal history of the build from in-situ thermal streams against the twin's prediction, estimates residual stress and distortion as the part grows, and recommends thermal compensation, scan-strategy or dwell changes before geometry leaves tolerance.

thermal.read_fieldtwin.compare_thermalprinter.set_dwell

agent.defect_inspect · shipped as Poroscan

Defect-and-Inspect

CT, vision and metrology inspection, grading, qualification evidence

Triages CT volumes for pores, cracks and lack of fusion with GPU 2D and 3D segmentation, fuses surface vision and metrology deviation, grades each part against the alloy and part acceptance criteria with a human sign-off checkpoint, and links every defect back to the in-situ signal and the recipe that produced it.

ct.segmentmetrology.comparepacket.assemble

agent.postprocess_heat

Postprocess-and-Heat

Stress relief, heat treatment, HIP, support removal

Holds stress-relief, heat-treatment and HIP recipes against the part's as-built state and the qualification spec, sequences support removal, and records each thermal cycle into the part genealogy so post-processing stops being the blind, reactive end of the floor.

oven.read_cycleoven.set_recipegenealogy.write_cycle

agent.yield_takt · shipped as Yieldex

Yield-and-Takt

Yield, build-failure risk, energy, fleet scheduling

Ingests build health, inspection results, recipes and cell state from the whole floor, predicts build-failure and qualification risk per job, and schedules machines, ovens, HIP, inspection cells and robots for yield, takt and energy with cuOpt, re-optimising after a delay or a quality hold.

cuopt.schedulerisk.predict_failuremes.read_orders

agent.robot_handling · shipped as Forgebot

Robot-and-Handling

Build-plate, powder, part handling and depowdering

Drives Isaac and Jetson robot cells that load and unload build plates, handle and sieve reactive powder, depowder parts and move them through heat treatment, HIP and support removal, with motion planned and checked in simulation and coordinated with live printer and oven state.

isaac.plan_graspcell.transferpowder.sieve

agent.quality_conf

Quality-and-Conformance

Right-first-time, non-conformance, build genealogy, qualification traceability

Maintains the immutable build genealogy from recipe to layer to inspection to disposition, runs non-conformance workflows, and blocks shipment when genealogy or defect evidence is incomplete, so every qualified part carries a record a certification engineer can audit.

genealogy.writencr.openpacket.sign

agent.am_knowledge

AM-Knowledge

Parameter sets, specs, AM standards, procedures, operator Q&A

Answers process and qualification questions from parameter sets, build and qualification specs, AM standards and quality procedures, with every answer cited to its source and scoped to the tenant, so engineers get grounded reasoning rather than a chat window.

rag.retrievestandards.citeprocedure.lookup

twin.parttwin · shipped as Parttwin

Part Twin

As-built part simulation and recipe optimisation before the build

Simulates the build for candidate orientation, support, nesting and parameter recipes: thermal fields, a melt-pool proxy, residual stress and distortion. Auto-optimises against density, geometry, takt and energy targets, stress-tests the winner against rare defect and distortion scenarios, and hands the validated recipe to Meltrix.

twin.simulatetwin.rank_recipesscenario.generate

agent.orchestrator

Floor Orchestrator

Planning, arbitration, human approval

Plans the build job, routes events to the right agent, arbitrates between agents competing for the same machine or cell, enforces the autonomy level and the OEM parameter envelope, and sends any qualification-impacting action to a named approver.

plan.composepolicy.evaluateapproval.request
Agent by agent

What each agent watches, moves and measures

No agent moves a parameter it has not declared. The metric column is what a pilot on that agent would measure against the floor's baseline.

agent.prep_parameter

Prep-and-Parameter

Proposes orientation, support, nesting and process-parameter recipes (laser power, speed, hatch, layer) for a build, scores them against density, geometry and takt targets with the part twin, and releases the winning recipe to the printer only inside the OEM-approved parameter envelope and after sign-off.

Part of the platform, included in the Floor tier.

slicer.read_buildtwin.rank_recipesprinter.release_recipe
Build preparation, orientation, supports, nesting, process parameters SUCCEEDED
Prep-and-Parameter: signals, moves, metric
WatchesPart geometry, alloy or polymer lot, machine state, approved parameter envelope, twin predictions
MovesOrientation, supports, nesting and parameter recipe, released to the printer on approval
MetricFirst metric on a pilot: first-time-right builds and coupon campaigns avoided
agent.melt_layer

Melt-and-Layer

Fuses coaxial and off-axis melt-pool video, layer and recoater images, photodiode and thermal streams at high frame rate, classifies and predicts porosity, lack of fusion, keyholing, spatter and recoat faults seconds ahead, and closes into bounded, operator-graduated laser and energy control to hold the pool stable.

Shipped as Meltrix.

holoscan.fuseprinter.trim_powergenealogy.tag_layer
In-situ melt-pool, layer and energy control SUCCEEDED
Melt-and-Layer: signals, moves, metric
WatchesMelt-pool video, layer and recoater images, photodiode and thermal streams, layer number
MovesLaser power, speed and focus inside the approved envelope; alerts and layer tags
MetricFirst metric: defect prediction accuracy against CT, then scrapped builds avoided
agent.recoat_powder

Recoat-and-Powder

Watches recoater images and bed uniformity every layer, detects streaks, short feeds and spreadability faults before the next layer is fused, tracks powder lot, reuse count and quality, and recommends or applies bounded recoat and dosing corrections.

Part of the platform, included in the Floor tier.

recoater.read_imageprinter.adjust_dosepowder.lot_record
Recoating, powder-bed uniformity, powder quality and reuse SUCCEEDED
Recoat-and-Powder: signals, moves, metric
WatchesRecoater camera, bed uniformity map, powder lot, sieve and reuse history, humidity
MovesRecoat speed and dosing inside limits; hold requests to the orchestrator
MetricFirst metric: recoat-linked build stops and scrapped layers
agent.thermal_distortion

Thermal-and-Distortion

Tracks the thermal history of the build from in-situ thermal streams against the twin's prediction, estimates residual stress and distortion as the part grows, and recommends thermal compensation, scan-strategy or dwell changes before geometry leaves tolerance.

Part of the platform, included in the Floor tier.

thermal.read_fieldtwin.compare_thermalprinter.set_dwell
Thermal fields, residual stress, distortion SUCCEEDED
Thermal-and-Distortion: signals, moves, metric
WatchesThermal camera and pyrometer streams, interlayer time, plate temperature, twin thermal prediction
MovesInterlayer dwell, scan-strategy order, compensation recommendations to Prep-and-Parameter
MetricFirst metric: geometry deviation at metrology against baseline
agent.defect_inspect

Defect-and-Inspect

Triages CT volumes for pores, cracks and lack of fusion with GPU 2D and 3D segmentation, fuses surface vision and metrology deviation, grades each part against the alloy and part acceptance criteria with a human sign-off checkpoint, and links every defect back to the in-situ signal and the recipe that produced it.

Shipped as Poroscan.

ct.segmentmetrology.comparepacket.assemble
CT, vision and metrology inspection, grading, qualification evidence SUCCEEDED
Defect-and-Inspect: signals, moves, metric
WatchesCT volumes, surface images, metrology scans, acceptance criteria, Meltrix build-health stream
MovesNothing on the printer. Accept, reject and rework suggestions, inspection routing, scan priority
MetricFirst metric: CT triage time and detection accuracy against the inspector baseline
agent.postprocess_heat

Postprocess-and-Heat

Holds stress-relief, heat-treatment and HIP recipes against the part's as-built state and the qualification spec, sequences support removal, and records each thermal cycle into the part genealogy so post-processing stops being the blind, reactive end of the floor.

Part of the platform, included in the Floor tier.

oven.read_cycleoven.set_recipegenealogy.write_cycle
Stress relief, heat treatment, HIP, support removal SUCCEEDED
Postprocess-and-Heat: signals, moves, metric
WatchesOven and HIP cycle data, part and alloy spec, as-built state from Meltrix and Parttwin
MovesOven and HIP recipe parameters inside the qualified window; support-removal sequence
MetricFirst metric: post-processing non-conformances and cycle time
agent.yield_takt

Yield-and-Takt

Ingests build health, inspection results, recipes and cell state from the whole floor, predicts build-failure and qualification risk per job, and schedules machines, ovens, HIP, inspection cells and robots for yield, takt and energy with cuOpt, re-optimising after a delay or a quality hold.

Shipped as Yieldex.

cuopt.schedulerisk.predict_failuremes.read_orders
Yield, build-failure risk, energy, fleet scheduling SUCCEEDED
Yield-and-Takt: signals, moves, metric
WatchesFleet state, build-health stream, inspection outcomes, order queue, oven and HIP queues, energy windows
MovesSchedule, machine assignment, inspection and post-processing sequence
MetricMeasured as build-failure rate, machine utilisation and energy per part against baseline
agent.robot_handling

Robot-and-Handling

Drives Isaac and Jetson robot cells that load and unload build plates, handle and sieve reactive powder, depowder parts and move them through heat treatment, HIP and support removal, with motion planned and checked in simulation and coordinated with live printer and oven state.

Shipped as Forgebot.

isaac.plan_graspcell.transferpowder.sieve
Build-plate, powder, part handling and depowdering SUCCEEDED
Robot-and-Handling: signals, moves, metric
WatchesBuild-complete signals, plate and part pose, powder container state, oven availability, safety zones
MovesPlate load and unload, depowdering, part transfer, powder handling
MetricMeasured as machine idle time between builds and manual handling hours removed
agent.quality_conf

Quality-and-Conformance

Maintains the immutable build genealogy from recipe to layer to inspection to disposition, runs non-conformance workflows, and blocks shipment when genealogy or defect evidence is incomplete, so every qualified part carries a record a certification engineer can audit.

Part of the platform, included in the Floor tier.

genealogy.writencr.openpacket.sign
Right-first-time, non-conformance, build genealogy, qualification traceability SUCCEEDED
Quality-and-Conformance: signals, moves, metric
WatchesEvery recipe release, layer tag, approval, inspection result and post-processing cycle on the build
MovesRelease, hold and non-conformance decisions; shipment block when evidence is incomplete
MetricMeasured as genealogy completeness and non-conformance cycle time
agent.am_knowledge

AM-Knowledge

Answers process and qualification questions from parameter sets, build and qualification specs, AM standards and quality procedures, with every answer cited to its source and scoped to the tenant, so engineers get grounded reasoning rather than a chat window.

Part of the platform, included in the Floor tier.

rag.retrievestandards.citeprocedure.lookup
Parameter sets, specs, AM standards, procedures, operator Q&A SUCCEEDED
AM-Knowledge: signals, moves, metric
WatchesParameter sets, AS9100, ISO 13485, ASTM F42 and ISO 52900 documents, non-conformance history
MovesNothing. It informs operators and the other agents and cites what it used
MetricMeasured as citation coverage on every answer
twin.parttwin

Part Twin

Simulates the build for candidate orientation, support, nesting and parameter recipes: thermal fields, a melt-pool proxy, residual stress and distortion. Auto-optimises against density, geometry, takt and energy targets, stress-tests the winner against rare defect and distortion scenarios, and hands the validated recipe to Meltrix.

Shipped as Parttwin.

twin.simulatetwin.rank_recipesscenario.generate
As-built part simulation and recipe optimisation before the build SUCCEEDED

As-built part simulation and recipe optimisation before the build.

agent.orchestrator

Floor Orchestrator

Plans the build job, routes events to the right agent, arbitrates between agents competing for the same machine or cell, enforces the autonomy level and the OEM parameter envelope, and sends any qualification-impacting action to a named approver.

Part of the platform, included in the Floor tier.

plan.composepolicy.evaluateapproval.request
Planning, arbitration, human approval SUCCEEDED

Planning, arbitration, human approval.

Run timeline · scenario

One scenario run, step by step

Every Powdrix run is an ordered, inspectable sequence. This is run_7c2e4b on PBF-2 · Ti-6Al-4V bracket family, 24-part build. Scenario completes in 10 steps with one approval gate. Twin recipe #14 selected, one lack-of-fusion precursor at layer 1,872 cleared by a bounded power trim, CT triage in 11 minutes, 23 of 24 parts released with full genealogy and one held for review.

run_7c2e4b PBF-2 · Ti-6Al-4V bracket family, 24-part build illustrative scenario SUCCEEDED
  1. 01 ingest.job SUCCEEDED 0.7 s

    Floor Orchestrator · Pulled the job from AM MES: 24 Ti-6Al-4V brackets, 60 µm layers, 3,140 layers, density target 99.7%, geometry tolerance ±0.10 mm, AS9100 evidence packet required. Locked the OEM-approved parameter envelope for PBF-2.

  2. 02 twin.simulate SUCCEEDED 6 m 12 s

    Part Twin · Simulated 40 candidate orientation, support and parameter recipes for thermal field, residual stress and distortion, ranked on predicted density, geometry deviation, takt and energy.

  3. 03 prep.release SUCCEEDED 48 s

    Prep-and-Parameter · Recipe #14 released: 32° orientation, lattice supports on two faces, laser power and speed inside the qualified envelope. Predicted distortion 0.06 mm, predicted density 99.8%.

  4. 04 recoat.check SUCCEEDED continuous

    Recoat-and-Powder · Recoater image checked every layer. Powder lot L-2291 on its third reuse, humidity in window. One short-feed at layer 412 corrected with a dosing increase before the next layer fused.

  5. 05 melt.stabilise SUCCEEDED continuous

    Melt-and-Layer · Melt-pool, photodiode and thermal streams fused at 120 fps. Lack-of-fusion risk rose to 0.61 on part 17 at layer 1,872 and was cleared by a 3% power trim inside the envelope within four layers.

  6. 06 thermal.compensate SUCCEEDED 2 m 04 s

    Thermal-and-Distortion · Interlayer temperature on the tall brackets ran 14 °C above the twin's prediction from layer 2,300. A 1.5 s interlayer dwell held the thermal field inside the predicted band through the top section.

  7. 07 approve.human APPROVAL 4 m 30 s

    Floor Orchestrator · Melt-and-Layer proposed a hatch-spacing change on part 17 beyond the qualified envelope. Held for the AM process engineer, who declined it and chose to flag part 17 for CT priority instead. Decision logged.

  8. 08 handle.depowder SUCCEEDED 38 m

    Robot-and-Handling · Plate unloaded by the handling cell on build complete, depowdered, and parts transferred to stress relief. Powder returned to the sieve with lot L-2291 reuse count incremented in genealogy.

  9. 09 inspect.ct SUCCEEDED 11 m

    Defect-and-Inspect · 24 CT volumes segmented for pores, cracks and lack of fusion in 11 minutes. 23 parts graded accept against the acceptance criteria. Part 17 shows a 180 µm indication at layer 1,870 and is routed to the inspector for disposition.

  10. 10 release.qualify SUCCEEDED 1 m 20 s

    Quality-and-Conformance · 23 parts released with the full genealogy: recipe, layer tags, melt-pool events, thermal trace, approvals, post-processing cycles and CT result. Part 17 held with a non-conformance opened. Evidence packet assembled for sign-off.

Multi-agent handoff

Agents negotiate, they do not collide

Two agents will want the same machine or the same parameter. The orchestrator arbitrates on the job goal and the qualification rules, not on who asked first, and the handoff is logged like any other step.

Contested move: PBF-2 laser power on part 17

  1. 01yield_takt.request(speed +4%) SUCCEEDED0.2 s

    Yield-and-Takt wants to raise scan speed on the build to recover takt lost to the recoat correction at layer 412.

  2. 02melt_layer.request(power −3%) SUCCEEDED0.2 s

    Melt-and-Layer wants to trim laser power on part 17 to clear a lack-of-fusion precursor at layer 1,872.

  3. 03orchestrator.arbitrate SUCCEEDED0.4 s

    Lack-of-fusion risk 0.61 on a qualified part outranks 1.1 hours of takt under the job goal. Melt-and-Layer wins the parameter for four layers; the speed change is deferred.

  4. 04yield_takt.handoff(rescheduled) SUCCEEDED1.1 s

    Yield-and-Takt recovers the takt by re-sequencing the oven queue with cuOpt instead. Density prediction holds at 99.8%.

Illustrative scenario.

Arbitration rules

  • Job goal first. Every request is scored against the declared density, geometry and qualification targets, not the requesting agent's local objective.
  • Qualification outranks takt. A pore in a qualified part costs more than an hour of machine time.
  • The envelope is not negotiable. No arbitration outcome can leave the OEM-approved parameter envelope; that path always goes to a human.
  • Time-boxed ownership. An agent holds a contested parameter for a bounded number of layers, then must re-justify.
  • Everything is logged. The losing request, the score and the reason all appear in the build genealogy.
  • Deadlocks escalate to a human rather than resolving by timeout.

See the agent roster

Autonomy

Four levels, set per agent and per machine

A qualified floor does not go from manual to unattended in one step. Powdrix makes the level explicit, auditable and reversible at any time, and human approval for anything that affects qualification stays mandatory until validation exists. The first release runs shadow mode, then assist mode, then graduated autonomy.

Autonomy levels and the human role at each
LevelWhat the agent doesWhat the human doesWhere it sits in a pilot
L1 · AdvisoryPredicts and recommends, with citations; nothing is writtenEnters every change manuallyShadow mode on one machine, first weeks of a pilot
L2 · SupervisedProposes a parameter, thermal or inspection move; it executes on approvalApproves each move in the review consoleAssist mode once accuracy is measured against CT
L3 · BoundedWrites inside the OEM-approved parameter envelope and per-tag limitsApproves anything that leaves the envelope or affects qualificationGraduated autonomy on low-risk loops, if the floor is satisfied
L4 · UnattendedRuns the envelope without promptingSets the envelope; reviews the shift recordPlanned; only after validation on the part family
Guardrails

An agent that can trim laser power needs a leash

Powdrix writes to production printers that make qualified parts. Every capability is scoped to the qualified envelope, every write is policy-checked, and every action is written to an append-only audit log the floor owns.

  • Bounded action space. Each agent writes only to an explicit allow-list of printer, oven and cell parameters, inside the OEM-approved envelope and per-tag rate and magnitude limits.
  • Policy engine before every write. Autonomy level, part family, qualification status, machine state and operator presence are evaluated before a parameter moves.
  • Human approval for qualification-impacting actions. Anything that leaves the qualified envelope, changes a validated recipe or affects a part's disposition waits for a named approver. This stays mandatory until validation exists.
  • Immutable audit log. Append-only and exportable. Every action records model version, data lineage, source signals, predicted impact, allowed range, approval status and qualification impact.
  • Hard fallback. Loss of the edge node, the network or a model returns control to the printer's own controller and the standing recipe.
  • Tenant and IP isolation. Parameter sets, part geometry, defect libraries and process memory never cross a customer boundary. On-prem deployment is available for sensitive producers.

Compliance posture

Compliance and certification status
ItemScopeStatus
SOC 2 Type ICloud control plane QUEUED Planned, first 6 months
SOC 2 Type IICloud control plane QUEUED Planned, 6 to 12 months
SSO and RBACConsole and approvals QUEUED Planned, 6 to 12 months
AS9100 and ISO 13485 assurance packagesEvidence packets and model change control QUEUED Planned, 6 to 12 months; not a certification claim
Standards-cited reasoningAS9100, ISO 13485, ASTM F42, ISO 52900 RUNNING In the design, every answer cites its source
Per-tenant isolation and on-prem optionRecipes, geometry, models, memory RUNNING In the design from day one

Read the security overview

FAQ

Straight answers

The questions operations directors, process engineers and certification engineers ask in the first meeting.

Only inside the OEM-approved parameter envelope, only on an explicit allow-list with per-tag limits, and only at the autonomy level your floor has set. Level 1 is advisory: Powdrix predicts and recommends and a human enters everything. Design partners start there, in shadow mode. Anything that leaves the envelope or affects qualification always waits for a named approver.

Get started

Put one agent on one machine

Start with one printer or one inspection cell and one measurable baseline. A paid pilot on melt-pool and build-failure prediction, CT porosity triage or pre-build recipe optimisation shows the number before you commit further.