Menu
Siemens ICXSolutionsHow we deliverResultsInsightsAboutEventsAcademyCareersContact
PTENES
Contact
Delivery

People and agents, one team: proof at every step and a focus on value, not hours.

Mason, our proprietary AI technology, used under exclusivity with AGENIUS, does the heavy lifting on Siemens ICX; our people decide, validate and answer for the outcome, together with the client.

How we deliver

Cognitive teams: people and AI agents in one team.

Mason, powered by AGENIUS, abstracts the complexity of adopting Siemens ICX and cuts implementation costs and timelines dramatically. Agents do the heavy lifting; our people decide, validate and own the result.

01Frame
Mason

Reads the available systems and data.

People

Agree outcome, baseline and acceptance criteria.

decision
02Accelerate
Mason

Maps the legacy and proposes design, ontology and phases.

People

Challenge the proposal and decide the design.

decision
03Refine and build
Mason

Builds on ICX and generates the tests.

People

Iterate with the business and review what goes in.

decision
04Prove and evolve
Mason

Runs the tests and monitors operations.

People

Accept on evidence and answer for the result.

decision
Mason investigating a legacy system: each light is an agent reading a file. Excerpt from the technical session demo at Siemens Transform 2026.
Control

Every critical decision is traceable and validated by people. Guardrails are deterministic: what fails a check does not go in.

Contract

Engage for value, not hours: hybrid discovery, phases measured by the agreed outcome, and continuous evolution.

Engagement

How we engage with clients.

A four-step engagement model built for the hybrid team: agents compress the time, people secure the result.

  1. 0Hybrid discoveryWeeks, not months

    Mason reads systems, data and documents. With the sponsor we agree the problem, the baseline, the acceptance criteria and the business case.

    Output Business case, legacy map, priced phase plan
  2. 1First bounded phase1 to 2-week cycles

    One critical decision, end to end: context, prediction and agentic application. Each cycle ends with something the client can try, a gate and evidence.

    Output Increment in production, evidence pack, acceptance
  3. 2ScalePhase by phase

    The ontology and agents from the first phase are reused. The squad becomes the standard unit, and each new use case starts further ahead.

    Output Use-case roadmap, reusable patterns
  4. 3Run and evolveOngoing

    AgentOps runs and observes. A periodic value review with the sponsor, and a decision on the next problem to solve.

    Output Monitoring, value review, next use case
Effectiveness

Doing the right thing.

  • Requirements start from evidence: the client corrects a draft instead of describing everything from memory.
  • Baseline and acceptance criteria set before building.
  • The client tries each increment and decides; nothing reaches production without acceptance.
  • Business context lives in a governed ontology, not in a few people’s heads.
Efficiency

Doing it right, with less.

  • Agents do the heavy lifting: investigate, generate, test, document.
  • Senior people spend their time where it counts: business rules, exceptions, integration and decisions.
  • Automated gates replace manual reviews; governance happens while building.
  • Small squads and reused patterns: each client starts further ahead than the last.
= ROI

Efficiency compresses the time; effectiveness secures the result. ROI = time × results.

What we ask of the client
  • SponsorStates the intent, names an owner and accepts each phase.
  • Process ownersCorrect the map Mason produced, with the people who run the process.
  • Key usersTry each increment and decide.
  • IT and securityOpen access and validate integration, data and compliance.
Cadence and governance
  • Every cycleIncrement demo, quality gate and updated evidence pack.
  • Every phaseFormal acceptance and value review against the baseline.
  • Steering committeeThe sponsor and S4 leadership review priorities, risks and the next use case.
  • ContractHybrid discovery, phases with a price and evidence and, if you wish, a performance component.
The squad

Anatomy of a cognitive squad.

Eight roles around one result, agents in the loop, Siemens underneath.

Equipe cognitiva: oito papéis humanos em volta do Mason e dos seus agentesMasonagentes executamagents executelos agentes ejecutanSucesso do ClienteCustomer SuccessÉxito del ClienteForward-DeployedFluxo da SquadSquad FlowFlujo del SquadSoluções ICXICX SolutionsSoluciones ICXDados e GrafosData & GraphsDatos y GrafosProduto + IAProduct + AIProducto + IAQualidade e EvidênciaQuality & EvidenceCalidad y EvidenciaPlataforma e AgentOpsPlatform & AgentOpsPlataforma y AgentOpsPessoas decidem, questionam e respondem pelo resultadoPeople decide, challenge and answer for the resultLas personas deciden, cuestionan y responden por el resultado
  • Client-facingCustomer Success & Innovation Leader, Forward-Deployed Engineer, Squad Flow Lead.They own the relationship, the requirements and the cadence.
  • Data to decisionICX Solution Engineer, Data Scientist, AI-Augmented Product Engineer.They design, ground and build on ICX.
  • Proof and runQuality & Evidence Engineer, AI Platform & AgentOps Engineer.They prove it works and keep it working.
  • Mason agentsInvestigate, design, build, verify, operate.They do the work in between; people decide and accept.

Small by design. A squad is three to five people. One person often holds two roles, and the shape changes by phase.

See the profiles in detail →
Who does what

Agents do the work. People own the outcome.

Mason appears as R or S; the A column is always a person.

AAccountableRDoes the workSSupports with draftsCConsultedIInformed
ActivityCS&IFDEFlowICX SEData Sci.Product Eng.QualityAgentOpsMason
Business case and baselineARICCIIIS
Investigate legacy, data and documentsIAICCCIIR
Iterative requirements with the clientCACCCCCIS
Context model: ontology and graphICICAIIIR
Solution design and phase planICCACCCCR
Build increments on Mendix / ICXIICCCACIR
Models and agents in AI StudioIIICACCCR
Verification and evidence packICCCCCAIR
Release, operate, observeIICCICCAR
Value review and next use caseARCCCIIIS
Patterns fed back to AGENIUSICIACCCCR

CS&I: Customer Success & Innovation Leader · FDE: Forward-Deployed Engineer · Flow: Squad Flow Lead · ICX SE: Siemens ICX Solution Engineer

NextResults →Talk to our team