Crucible

Verification-first deployment of AI agents: every action stress-tested in a world model before it touches real infrastructure.

Built for the systems that cannot afford an experiment: mobile networks and semiconductor fabs.

Telecom networks Energy, slice assurance, rApp conflicts and customer operations, orchestrating the features already in your vendors' equipment. E-RAN and the telecom agents → Semiconductor fabs Real-time dispatch: what next, where next and when, for every tool event. The policy is proven optimal on a certified model of the lot flow and stress-tested on the full-fab world model before it takes the seat. Fab Dispatch →

How It Works

Model the world

Crucible builds a calibrated world model of your system, a radio cluster or a whole fab: an AI-native simulation faithful enough to stand in for reality.

Stress-test the agents

Optimization agents are trained and proven against the model, never by experimenting on live infrastructure.

Gate every action

Actions release only after verification against your guardrails, and roll back automatically if reality deviates from prediction.

The same engine in both industries

Telecom networkSemiconductor fab
The world model a radio cluster: traffic, power, interference and quality of service, calibrated on your performance counters a full fab: lots, tools, batches, setups, time limits between steps and tool downs, calibrated on your MES history
What the agent decides which carriers sleep and when, which slice gets capacity, which rApp wins a conflict, which offer a subscriber receives which lot, which tool, which chamber, now or wait; batch composition, start time and setup sequence
The guardrails availability, latency and coverage: the quality-of-service envelope you define tool qualification, holds, time limits and carrier state: the fab's hard constraints, validated by the MES on every instruction
Proof before production recommendation mode on selected sites, then closed loop a policy proven optimal on the certified lot-flow core, full-fab simulation, then shadow mode on one area, then the dispatch seat
If reality deviates automatic rollback to the previous configuration kill switch back to the fab's native rules

Telecom Networks

E-RAN applies Crucible to radio access networks, which consume most of a telecom operator's energy budget. It orchestrates the energy-saving features already in your vendors' equipment, with zero hardware changes while safeguarding quality of service.

10–15%energy OpEx reduction
$450Ksaved per year per 1,000 sites
15,000tonnes of CO₂ avoided per 10,000 sites

Nokia · Ericsson · Huawei · Open RAN

In operation

24.6%energy saved over the simulated day (118.8 to 89.53 kWh)
0.015%call drop rate, unchanged from baseline
6 kW 0 00:00 24:00 always-on baseline with Crucible policy

one simulated day on a 21-cell cluster in the Crucible world model; cluster power draw, hourly means

00:30carrier sleep engaged, first of the night (cell at 2% load) verified: within QoS guardrailsreleased
05:15carriers woken ahead of the morning ramp demand forecast risingreleased
deeper sleep policy, 31% saving latency 7 to 22 ms in the world modelrejected, never deployed

logged from the same simulated day; seed 42, 40 Mbps/cell demand

Beyond energy, on the same world model


Semiconductor Fabs

Fab Dispatch applies Crucible to the decision layer of a wafer fab. It is the real-time dispatcher: for every tool event it answers what next, where next and when, as concrete instructions (lot to tool to chamber to load port, batch composition, start time, and deliberate, bounded waits). Its dispatch policy is proven optimal on a certified model of the fab's lot flow: it attains the machine-checked pace floor exactly, and a certified switching envelope benches any policy that leaves it. Each instruction goes to your MES, which validates it against the fab's hard constraints and executes it. The MES stays the system of record; on the toolsets Crucible runs, the native rules step back to a kill switch.

$120Ma year per 1% of throughput at the bottleneck
$33Mof work in progress released per day of cycle time
0Q-time violations (the maximum time allowed between two steps): a guardrail, not a target

orders of magnitude for a 100,000-wafer-per-month fab at $10,000 per wafer

Siemens Opcenter · Critical Manufacturing · Applied SmartFactory · in-house MES

In operation

0regret: Crucible's policy (π*) drains 84 lots in exactly 21,795 ticks, the machine-checked floor 255N + 375 that no policy can beat
+84%cycle time under the wrong dispatch rule (SPT) on the same fab; the best never-idle rule finishes 80 ticks (+0.4%) behind the optimum
ticks to complete all 84 lots Crucible π* the certified floor 255N + 375, proven SRPT / FIFO rule +80 ticks (+0.4%) SPT rule +18,305 ticks (+84%) LBFS + least advanced +18,611 ticks (+85%) rogue policy benched by the envelope certified floor, 21,795

ticks to complete all 84 lots on the certified Mini-Fab lot-flow core (carved clock, 1 tick = 1 minute); the darker segment is each rule's regret beyond the certified floor 255N + 375; the rogue policy never runs, the envelope benches it

tick 1first lots admitted to diffusion back-to-back on one furnace; the second furnace held 369 ticks for the re-entrant pass with a qualified lot waiting verified: inside the certified envelopereleased
tick 21,406furnace held 133 ticks with one lot waiting: the deliberate idle that keeps the last handover seamless verified: bounded wait, buys the final 80 ticksreleased
tick 21,795last lot arrives at lithography: exactly the certified floor 255N + 375 theorem: no policy arrives earliercertified
work-alternating rogue policy, +85% cycle time in the world model (40,406 vs 21,795 ticks) benched by the certified switching envelope after 8 named refusalsrejected, plant untouched

logged from the certified lot-flow core, 84 lots; the floor is machine-checked in Lean and the envelope certificates re-check from bytes. On the full fab (batching, setups, seeded failures, outside the certificate) the same engine re-plans daily and runs 12% shorter cycle time than FIFO over a simulated week.

Not a scheduler on top of your rule engine

A scheduler on top

publishes a list into the fab's rule engine every few minutes. The engine takes the first scheduled lot that happens to be at the tool and falls back to its own rules otherwise. Every layer between the list and the tool dilutes the result, which is why a scheduler on top cannot promise its own performance.

Crucible in the dispatch seat

answers each tool event in well under a second from the current plan and re-plans the toolset every few minutes. Every wait is deliberate, bounded and explained on the operator screen: "batch forming, 2 of 3", "hot lot arriving in 3 min". The measure is execution conformance: the share of tool starts that were Crucible's instruction, as issued.

Where it sits

Planning and ERPwafer starts, due dates, priority classes
Crucible Fab Dispatchwhat next, where next, when; every instruction proven in the world model
MESvalidates every instruction, stays the system of record
MCS and AMHStransport orders, vehicles, stockers
ToolsSECS/GEM, untouched

The MES holds no sequencing logic and Crucible holds no recipes. If Crucible stops, the fab keeps dispatching on its native rules.

Where a pilot starts

Control scopeone area; the data scope is the whole fab: work in progress, tool states and time-limit clocks fab-wide
First areathe wet-clean and diffusion pair: batch fill, furnace start timing and Q-time gating are the most valuable and the most local decisions
Thenimplant, where setup sequencing is the lever; lithography once the world model has earned the fab's trust, because that is where the fab-level money is
Proofa full-fab world model calibrated on your history, then shadow mode: instructions computed, not executed, and compared with what the fab actually did
Do no harmline yield, rework and scrap are watched as safety KPIs; recipes and process settings are never touched

From First Look to Full Scale

Assessmentpotential quantified in weeks, from exported counters or MES history World modelcalibrated to your network or your fab Pilotrecommendation mode on selected sites, or shadow mode on one area Scaleclosed loop, on your terms

Security and Compliance

ISO 27001certification in progress UK GDPRprivacy policy and DPA SOC 2mapped to the same controls

Crucible deploys as containers in your private cloud, connected or air-gapped. In a network it reads equipment-level performance and energy counters and never subscriber data. In a fab it reads lot, tool and route events from the MES: process times and constraints, never the recipe programs themselves. Humanis AI Ltd operates an ISMS built to ISO 27001, and a data processing agreement is available for every engagement.


Request an Assessment

Tell us about your network or your fab and we'll come back with a site-specific projection. No commitment.