Principal Batch Performance Consulting Lead — Processing Window Capacity

The mandate

Selected scheduled processing jobs in a retail and healthcare technology platform are approaching the end of their permitted operating windows. Teams estimate future completion by scaling last month's run time with record count, although job stages, data shape and overlapping work consume capacity differently. The consulting principal will deliver a Processing Window Capacity Study for six agreed jobs, showing the measured constraints and growth conditions that their owners can use to plan before late completion becomes routine.

Work begins on 26 October 2026 for four months, reserving three working days weekly. The first artifact, due on 30 November 2026, is a job-stage and workload diagnostic with an approved benchmark population. On 18 January 2027, the second milestone supplies measured capacity cases and competing-work experiments. The final milestone on 26 February 2027 contains internally replayed benchmarks, an operating-window recommendation file and maintenance instructions covering the completed study.

The analysis separates fixed setup time, volume-sensitive work and stages constrained by different resources. A job may finish quickly at an average record size but slow disproportionately when a larger population contains more expensive cases. Parallel execution may help one stage while delaying another job needed in the same window. The study must document those interactions and avoid turning an observed benchmark into a promise of linear future growth or a recommendation to add capacity before the actual limiting work is understood.

Joint acceptance rests with the Platform Engineering Head and Processing Service Owner. Internal engineers must reproduce the agreed baseline and two changed workload cases, explain material differences and recover the recommendation from the recorded method. Each material operating-window conclusion must identify the tested conditions, the stage that constrains completion and any dependency still unverified. Acceptance is possible with a visible limitation; it is not possible where the result depends on the consultant's undocumented interpretation of a run.

The sponsor provides approved workload extracts or synthetic equivalents, job histories, environment access, operating deadlines and three engineers with protected experiment time. Payments are 30% on diagnostic acceptance, 35% on the measured cases and 35% on final replay and transfer. Scheduling-platform implementation, functional data-correction rules and clinical or customer policy changes are excluded. New jobs, production migration or extended operating support require a signed scope and fee change; software and service owners retain implementation and operating responsibility.

What you will own

• Diagnose the six selected jobs through stage timing, workload shape and operating dependencies, establishing what the initial benchmark must represent before any future completion estimate is accepted.
• Define reproducible benchmark populations that vary meaningful record and work characteristics, distinguishing test convenience from the operating conditions relevant to the permitted processing window.
• Measure fixed and volume-sensitive stages through controlled experiments, testing where simple run-time extrapolation fails and recording the competing explanations considered before accepting a capacity conclusion.
• Examine overlapping job scenarios for their shared-resource effects, showing which scheduling assumptions remain plausible and which need an owner decision rather than prescribing a new scheduling system.
• Produce an operating-window recommendation file with measured envelopes and uncertainty, linking each proposed action to its underlying constraint and the evidence required before wider deployment.
• Facilitate internal benchmark replay and changed-case analysis, correcting instructions that prevent sponsor engineers from reproducing the findings or understanding why a result differs.
• Transfer benchmark maintenance and remeasurement triggers to job owners, identifying data-growth or stage changes that invalidate the earlier conclusion while keeping implementation outside the accepted project.

Candidate qualifications

• Bring seven or more years in performance architecture or engineering, with personal analysis of planned processing, workflow capacity or comparable sustained workloads. Describe how you identified the work that constrained completion and tested a growth assumption instead of accepting a proportional forecast. Relevant evidence may come from different processing patterns, but it must show an explicit measurement boundary and a recommendation traceable to observed behaviour.
• Demonstrate controlled experimental reasoning around data shape, stage duration, concurrency and shared resources. Explain the variables you would hold steady, the ones you would deliberately change and the observations needed to separate a stage bottleneck from interference by other work. No particular batch framework, programming language or infrastructure vendor is mandatory; you must work effectively with the sponsor's current engineering and measurement facilities.
• Show the ability to create a practical capacity study with stated conditions and usable confidence limits. Provide a finding that could not safely be generalised beyond the tested population, how you communicated that boundary and what additional experiment would have justified a broader recommendation. Retail or healthcare technology context is relevant, with functional and clinical policy decisions retained by their qualified owners.
• Establish principal specialist delivery through documented analysis, engineering review or method transfer that another practitioner could replay. Reserve three days weekly across the four-month term and identify the inputs needed from internal owners. You should be able to define acceptance before analysis, manage added jobs through commercial change control and leave maintenance instructions that do not require continuing consultant presence to interpret a changed run.

Application

Applications for this mandate are received in one way only: through the India Board Terminal's application process. It is automated end to end. Your Executive Passport travels to the mandate holder in its confidential form, your answers to the three questions below are read before anything else in your file, and every stage that follows is recorded on your applications page.

There is no address to write to and no intermediary to call. The mandate holder reads what the Terminal delivers and nothing else, which is what keeps the process the same for every applicant and keeps your name out of it until you release it. Applications close on 13 October 2026. Mandate reference CVU-CON-2026-IND-086.