Rugby AI Lead

The role will combine modern data science with generative and agentic AI to build longitudinal player models, identify emerging talent and changing trajectories, automate intelligence workflows and develop tools that strengthen human judgement and decision-making.

A central responsibility will be developing a secure agentic AI capability and digital workforce for Talent Insight & Management. This will move beyond individual AI tools towards interconnected agents capable of undertaking appropriate elements of research, monitoring, data gathering, video and scouting workflows, player tracking, quality assurance, analysis and intelligence preparation, while maintaining appropriate human oversight.

The role will continually explore and apply emerging approaches across AI, machine learning, predictive modelling and decision science, translating technological advances into practical applications capable of creating sustainable competitive advantage for Welsh rugby.

Salary: Competitive

• Strong practical experience applying data science, statistical modelling and machine learning to complex real-world problems.

• Advanced proficiency in Python and/or R, SQL and modern data-analysis and machine-learning environments.

• Degree, postgraduate qualification or equivalent professional expertise in data science, computer science, statistics, mathematics, artificial intelligence, machine learning or another relevant quantitative discipline.

• Strong practical experience applying data science, statistical modelling and machine learning to complex real-world problems.

• Advanced proficiency in Python and/or R, SQL and modern data-analysis and machine-learning environments.

• Demonstrable experience developing predictive models, including feature engineering, model selection, training, validation, tuning and ongoing performance monitoring.

• Strong understanding of statistical inference, uncertainty, hypothesis testing and the limitations of predictive models.

• Experience working with longitudinal, time-series or repeated-measures data and modelling change or progression over time.

• Practical understanding of modern generative AI and large language models, including their application to structured and unstructured information.

• Understanding of modern agentic AI architectures, including AI agents, tool use, retrieval, orchestration, workflow automation, multi-agent systems, memory and human-in-the-loop approaches.

• Ability to prototype and productionise AI-enabled workflows rather than simply use off-the-shelf AI applications.

• Experience working with cloud-based data and AI environments and integrating models into scalable analytics products or operational workflows.

• Strong capability in data engineering fundamentals, including data wrangling, integration, quality assurance and working with complex and imperfect datasets.

• Ability to translate technically complex analysis into clear, actionable intelligence for non-technical decision-makers.

• Strong curiosity and experimentation mindset, with the ability to rapidly evaluate emerging technologies and distinguish genuine capability from hype.

• Ability to work effectively across disciplines, combining quantitative evidence with expert human judgement and contextual information.

• Strong understanding of responsible AI, including privacy, security, explainability, bias, model governance and appropriate human oversight.

• A growth mindset, commitment to continuous learning

• Ability to communicate and work through the medium of Welsh is desirable.
  

• Develop longitudinal player intelligence, connecting historical and current information across identification, pathway, professional and international rugby to understand how players develop over time.

• Build player trajectory and potential models using statistical modelling, machine learning, time-series and other appropriate techniques to understand changing trajectories, probability of progression and future international potential.

• Develop predictive indicators of future performance, exploring the interaction between technical, tactical, physical, behavioural, maturation, opportunity, exposure and contextual factors.

• Create an agentic AI workforce capable of securely undertaking appropriate elements of data gathering, research, monitoring, video and scouting workflows, player tracking, quality assurance, reporting and intelligence preparation for human review.

• Develop human-in-the-loop AI-enabled player intelligence systems, including automated player profiles, trajectory visualisations, alerts, comparison tools, depth-chart and succession modelling, that augment expert judgement and enable coaches and talent staff to interrogate evidence, understand uncertainty and make better-informed decisions.

• Integrate structured and unstructured information, including performance data, physical information, video, scouting observations, coaching assessments, reports and other relevant evidence to create a richer understanding of each player.

• Develop and productionise advanced statistical and machine-learning models, ensuring models are appropriately validated, monitored and continually improved as new evidence becomes available.

• Translate complex analysis into decision-ready intelligence, creating intuitive visualisations, tools and outputs that allow coaches and senior leaders to understand player potential, uncertainty, development priorities, opportunities and risk.

• Establish rigorous AI and data standards, including data quality, model validation, explainability, privacy, security, bias assessment, auditability and responsible use of player information.

• Work trans-disciplinarily with rugby experts, combining data science and AI with coaching judgement, scouting expertise, talent science, performance science and human factors rather than treating analytics as a standalone discipline.

• Lead experimentation and applied research, rapidly testing emerging AI, machine-learning and decision-science approaches against genuine Welsh rugby problems and scaling those that demonstrate meaningful value.

• Maintain a future-facing technology horizon, developing relationships with leading researchers, technology organisations and practitioners and identifying emerging capabilities capable of materially improving how Welsh rugby identifies, understands and develops players.

Bydd y rôl yn cyfuno gwyddor data modern gyda deallusrwydd artiffisial cynhyrchiol ac agentig i adeiladu modelau chwaraewyr hir-dymor, gan adnabod talent newydd a newid trywydd, awtomeiddio llifoedd gwaith deallusrwydd a datblygu offer sy'n cryfhau barn a phenderfyniadau pobl.

Cyfrifoldebau canolog fydd datblygu gallu AI agentig diogel a gweithlu digidol ar gyfer Dealltwriaeth a Rheoli Talent. Bydd hyn yn symud y tu hwnt i offer AI unigol tuag at asiantau rhyng-gysylltiedig sy'n gallu cynnal elfennau priodol o ran ymchwil, monitro, casglu data, llifoedd gwaith fideo a sgowtio, olrhain chwaraewyr, sicrhau ansawdd, dadansoddi a pharatoi gwybodaeth, wrth gynnal goruchwyliaeth briodol gan bobl.

Bydd y rôl yn archwilio ac yn defnyddio dulliau newydd ym maes AI, dysgu peirianyddol, modelu rhagfynegol a gwyddor penderfyniadau yn barhaus, gan drosi datblygiadau technolegol yn rhaglenni ymarferol sy'n gallu creu mantais gystadleuol gynaliadwy i rygbi Cymru.

• Profiad ymarferol cryf o gymhwyso gwyddor data, modelu ystadegol a dysgu peirianyddol i broblemau cymhleth yn y byd go iawn.

• Hyfedredd uwch mewn Python a/neu R, SQL ac amgylcheddau dadansoddi data a dysgu peirianyddol.

• Gradd, cymhwyster ôl-raddedig neu arbenigedd proffesiynol cyfatebol mewn gwyddor data, cyfrifiadureg, ystadegau, mathemateg, deallusrwydd artiffisial, dysgu peirianyddol neu ddisgyblaeth feintiol berthnasol arall.

• Profiad ymarferol cryf o gymhwyso gwyddor data, modelu ystadegol a dysgu peirianyddol i broblemau cymhleth yn y byd go iawn.

• Hyfedredd uwch mewn Python a/neu R, SQL ac amgylcheddau dadansoddi data a dysgu peirianyddol.

• Profiad amlwg o ddatblygu modelau rhagfynegi, gan gynnwys peirianneg nodwedd, dethol model, hyfforddiant, dilysu, tiwnio a monitro perfformiad parhaus.

• Dealltwriaeth gref o gasgliad ystadegol, ansicrwydd, profi damcaniaeth a chyfyngiadau modelau rhagfynegi.

• Profiad o weithio gyda data hydredol, data cyfres amser neu a fesurir yn fynych a modelu newid neu ddilyniant dros amser.

• Dealltwriaeth ymarferol o AI cynhyrchiol modern a modelau iaith mawr, gan gynnwys eu cymhwyso i wybodaeth â strwythur a heb strwythur.

• Dealltwriaeth o bensaernïaeth AI agentig modern, gan gynnwys asiantau AI, defnydd offer, adalw, offeryniaeth, awtomatiaeth llif gwaith, systemau aml-asiant, dulliau cof a phobl yn rhan ohonynt.

• Gallu i greu prototeip a rhoi ar waith llifoedd gwaith wedi'u galluogi gan AI yn hytrach na dim ond defnyddio rhaglenni AI parod.

• Profiad o weithio gyda data cwmwl ac amgylcheddau AI ac integreiddio modelau yn gynnyrch dadansoddeg graddadwy neu lifoedd gwaith gweithredol.

• Gallu cryf yn hanfodion peirianneg data, gan gynnwys trosi data, integreiddio, sicrwydd ansawdd a gweithio gyda setiau data amherffaith a chymhleth.

• Gallu i drosi dadansoddiad cymhleth yn dechnegol yn wybodaeth glir, y gellir gweithredu arni ar gyfer penderfynwyr anhechnegol.

• Chwilfrydedd cryf a meddylfryd arbrofi, gyda'r gallu i werthuso'n gyflym dechnolegau ar ddod a gwahaniaethu rhwng gallu go iawn a heip.

• Gallu i weithio'n effeithiol ar draws disgyblaethau, gan gyfuno tystiolaeth feintiol gyda barn ddynol arbenigol a gwybodaeth gyd-destunol.

• Dealltwriaeth gref o AI cyfrifol, gan gynnwys preifatrwydd, diogelwch, natur i esbonio, rhagfarn, llywodraethu model ac arolygiaeth ddynol briodol.

• Meddylfryd twf, ymrwymiad i ddysgu parhaus

• Mae'r gallu i gyfathrebu a gweithio trwy gyfrwng y Gymraeg yn ddymunol.

• Datblygu deallusrwydd ar ei hyd am chwaraewyr, gan gysylltu gwybodaeth hanesyddol a chyfoes ar draws adnabod, llwybr, rygbi proffesiynol a rhyngwladol i ddeall sut mae chwaraewyr yn datblygu dros amser.

• Adeiladu modelau trywydd a photensial chwaraewyr gan ddefnyddio modelu ystadegol, dysgu peirianyddol, cyfresi amser a thechnegau priodol eraill i ddeall trywydd sy'n newid, tebygolrwydd dilyniant a photensial rhyngwladol yn y dyfodol.

• Datblygu dangosyddion rhagfynegol o berfformiad yn y dyfodol, gan archwilio'r rhyngweithio rhwng ffactorau technegol, tactegol, corfforol, ymddygiadol, aeddfedu, cyfle, amlygiad a chyd-destunol.

• Creu gweithlu AI agentig sy'n gallu ymgymryd yn ddiogel ag elfennau casglu data priodol, ymchwilio, monitro, llifoedd gwaith fideo a sgowtio, olrhain chwaraewyr, sicrwydd ansawdd, adrodd a pharatoi deallusrwydd i'w hadolygu gan bobl.

• Datblygu systemau deallusrwydd chwaraewr wedi'i alluogi gan AI gyda phobl yn rhan ohonynt, gan gynnwys proffiliau chwaraewr awtomataidd, delweddu trywydd, rhybuddion, offer cymharu, modelu siart dyfnder ac olyniaeth, sy'n ychwanegu barn arbenigol ac yn galluogi hyfforddwyr a staff talent i archwilio tystiolaeth, deall ansicrwydd a gwneud penderfyniadau mwy gwybodus.

• Integreiddio gwybodaeth â strwythur a heb strwythur, gan gynnwys data perfformiad, gwybodaeth ffisegol, fideo, arsylwadau sgowtio, asesiadau hyfforddi, adroddiadau a thystiolaeth arall berthnasol i greu dealltwriaeth gyfoethocach o bob chwaraewr.

• Datblygu a chynhyrchu modelau ystadegol a dysgu peirianyddol uwch, gan sicrhau bod y modelau yn cael eu dilysu'n iawn, eu monitro ac yn cael eu gwella'n barhaus wrth i dystiolaeth newydd ddod ar gael.

• Trosi dadansoddiad cymhleth yn ddeallusrwydd parod am benderfyniad, gan greu delweddu, offer ac allbynnau sy'n hawdd eu deall sy'n caniatáu i hyfforddwyr ac arweinwyr uwch ddeall potensial chwaraewr, ansicrwydd, blaenoriaethau datblygu, cyfleoedd a risg.

• Sefydlu safonau llym ar gyfer AI a data, gan gynnwys ansawdd data, dilysu model, gallu esbonio, preifatrwydd, diogelwch, asesu rhagfarn, archwiliadwyedd a defnydd cyfrifol ar wybodaeth chwaraewr.

• Gweithio ar draws disgyblaethau gydag arbenigwyr rygbi, gan gyfuno gwyddor data ac AI gyda barn hyfforddi, arbenigedd sgowtio, gwyddor dawn, gwyddor perfformiad a ffactorau dynol yn hytrach na thrin dadansoddeg fel disgyblaeth ar ei phen ei hun.

• Arwain arbrofi ac ymchwil cymhwysol, gan brofi dulliau deallusrwydd artiffisial newydd, dysgu peirianyddol a gwyddor penderfyniadau yn gyflym yn erbyn problemau go iawn rygbi Cymru a graddio'r rheini sy'n dangos gwerth ystyrlon.

• Cynnal gorwel dechnoleg sy'n edrych tua'r dyfodol, gan ddatblygu perthnasoedd gydag ymchwilwyr blaenllaw, sefydliadau technoleg ac ymarferwyr ac adnabod galluoedd newydd sy'n gallu gwella'n sylweddol sut mae rygbi Cymru yn adnabod, yn deall ac yn datblygu chwaraewyr.