Cj O'Neill

• • Led the delivery of Machine Learning outputs in preparation for a vendor Bake-off at German Tier 1 Retailer, co-ordinating cross-functional activities spanning Test, Deployment, Commercial and third party Hardware Partner dependencies (Diebold Nixdorf). My client was subsequently selected as a preferred vendor and awarded a rollout of their solution across 100 stores in the US market.

• • Conducted a comprehensive ML governance audit of the client's Front Facing Camera Shrink Detection solution - spanning data provenance, annotation methodology, evaluation integrity, model architecture, and regulatory compliance - producing a formal ten-entry RAID log presented at C-Suite level. Findings included data leakage in model evaluation, train/test domain mismatch, class imbalance and misrepresentation risk, and EU AI Act compliance gaps - formally documented and escalated to the Senior Leadership Team prior to any commercial commitment being made.

• • Built a best-practice Video Annotation Guide from scratch, replacing a previous version which was driving an annotation accuracy rate of just 50%. Introduced the Confirmability Rule as a governing principle for each Use Case, enabling the 3 rd party annotation partner to apply labelling criteria consistently. This has lifted annotation accuracy to over 90% and established the annotation quality foundation required to develop temporal models.

• • Identified the absence of a consistent model card format across SeeChange's AI solutions. Subsequently designed a standardised Model Card Template and built a Streamlit-based Model Card Generator tool

◦ - deployed internally to automate compliance documentation in line with Article 11 of the EU AI Act. The potential commercial and legal liability exposure in the EU retail market was quantified and escalated to the Senior Leadership Team, accepted as a formal remediation priority ahead of the client's planned rollouts.

Parachuted into a complex, multi-geography data programme for Walgreens Boots Alliance to stabilise delivery and drive the mapping of No7 Beauty Company's Sales Transaction data to STEP Stibo - managing cross functional teams across the UK, Italy and India alongside strategic relationships with Tata Consultancy Services and Cognizant.

• • Successfully mapped 85% of No7 Beauty Company's Sales Transactions to the corresponding STEP Stibo data by the agreed target date - significantly enhancing Data Accuracy and Consistency and improving the quality of Analytics and Reporting available to Business End Users.

• • Collaborated with the client's RunOps team to diagnose and repair 6 broken third-party Sales Feeds, restoring the integrity of sales data sourced from the Databricks Platform and safeguarding downstream reporting accuracy.

• • Identified that the Programme Delivery stalled due to the absence of a prioritised, well-defined Backlog. Subsequently recommended and on-boarded a contract Business Analyst who established a refined, prioritised Backlog which met a specified 'Definition of Ready'. This directly contributed to the successful on-time delivery of the Data Mapping exercise.

Successfully delivered engagements across eCommerce, Product Information Management (PIM) and Master Data Management (MDM) at Estee Lauder Companies, Charlotte Tilbury and Nespresso.

• • Led the rollout of Klarna Payment methods across Estee Lauder companies top 5 UK brands by revenue prior to the 2022 'Peak' Trading period. Built and maintained the client relationship end-to-end, ensuring each brand received the necessary marketing materials and training ahead of launch. The successful rollout generated over £300k of incremental revenue cumulatively over the course of this trading period.

• • Implemented a Drupal Content Management System (CMS) to empower the Darphin UK Brand Team to create and update website content independent of Engineering support. The re-platformed and re designed website delivered an enhanced customer experience and improved cart conversion rates by 30%.

• • Owned the £2 million PIM Implementation Programme budget at Charlotte Tilbury with direct accountability to the Project Sponsor. Led the Internal QA effort which ensured the most critical build defects were resolved and re-tested ahead of Super User Demos and client-led End-to-End Testing. Produced brand-specific Training Decks to drive Business User adoption of the new PIM solution.

• • Stabilised a Semarchy MDM Programme at Nespresso that had fallen behind the planned delivery timeline. Brought Programme Costs under control and established a refined & prioritised backlog which met the 'Definition of Ready' to guide the required Engineering and Test effort.