Applied AI · Retail · FinTech
AI Product Leader
15+ years driving digital transformation across Retail, FinTech, and SaaS. Now building and deploying AI agents that solve real business problems.
I'm an AI Product Leader with 15+ years of experience driving end-to-end digital transformation across eCommerce, FinTech, ElderTech, and Utilities. I've led product strategy across iconic brands including The North Face, Vans, Timberland, and Williams-Sonoma, Inc..
I'm now focused on the intersection of AI and product management — building, evaluating, and shipping AI-powered systems that create measurable business outcomes. This portfolio showcases AI agents I designed and deployed across Retail, Retail SaaS, and FinTech domains.
Each agent demonstrates a core concept in applied AI — from RAG and guardrails to multi-step orchestration and multimodal analysis.
Agent 01
Grounded customer service agent that answers policy questions from a verified document — with guardrails that block off-topic queries and escalate fraud.
VIEW CASE STUDY →
Agent 02
Full RAG pipeline over a 100-product catalogue — answers questions by retrieving from Pinecone and showing which products were used.
VIEW CASE STUDY →
Agent 03
LLM-as-Judge pipeline — runs 20 benchmark questions against Agent 02, scores each on Accuracy, Completeness, and Grounding, and produces a pass-rate report.
VIEW CASE STUDY →
Agent 04
Support bot for Finsera digital banking — answers account questions, refuses financial advice, blocks competitor mentions, and escalates fraud immediately with 6 active guardrails.
VIEW CASE STUDY →
Agent 05
Generates on-brand marketing copy in multiple formats — email subject line, product description, social caption, and push notification — from a product name and key features.
VIEW CASE STUDY →
Agent 06
Accepts product images and returns structured descriptions, attributes, and quality flags — combining vision and language for automated catalogue enrichment.
VIEW CASE STUDY →
Agent 07
Hybrid 2-call agentic pipeline for Caden Electronics — intent extraction, zero-token JS order lookup, and contextual support responses with a live reasoning trace visible in the UI.
VIEW CASE STUDY →
Agent 08
Confidence scoring, HITL governance thresholds, and auto-approve / route / escalate logic — demonstrating how AI systems should decide when to act autonomously vs. defer to humans.
VIEW CASE STUDY →
Agent 09
Parses unstructured documents — PDFs, emails, forms — and outputs validated JSON conforming to a defined schema, with field-level confidence scores on every extracted value.
VIEW CASE STUDY →
Agent 10
Adversarially tests LLM systems for prompt injection vulnerabilities — identifying attack vectors, testing guardrail robustness, and suggesting mitigations.
VIEW CASE STUDY →
Agent 11
Recommends products based on customer profile, purchase history, and browsing context — with conditional reasoning and explainable justification for each recommendation.
VIEW CASE STUDY →
Agent 12
Transforms a brief product idea into a structured PRD — including problem statement, user stories, acceptance criteria, and success metrics — using AI to scaffold product thinking.
VIEW CASE STUDY →
Agent 13
Classifies sentiment, clusters themes, and ranks issues by frequency across raw customer reviews — outputting a prioritised product insight report with business-ready findings.
VIEW CASE STUDY →
Agent 14
Takes an AI use case description and outputs estimated efficiency gain, implementation complexity, risk level, and recommended next steps — translating AI capability into business language.
VIEW CASE STUDY →
Agent 15
Pastes a GitHub URL and returns a live briefing — stack, activity signal, issue themes, and health verdict — powered by four real MCP tool calls to the GitHub API.
VIEW CASE STUDY →
Agent 16
Six specialist agents — orchestrator, four channel writers, and a synthesiser — transform one piece of content into a consistent campaign pack across Email, LinkedIn, Twitter/X, and Product Hunt.
VIEW CASE STUDY →
Agent 17
Transforms raw bank transaction strings into clean, categorised, human-readable descriptions — merchant name resolved, category classified, channel identified, and confidence scored in a single pass.
VIEW CASE STUDY →
Agent 18
Describe your marketing creative and get a precise, opinionated resizing brief for every format — crop anchors, safe zones, copy rules, and platform restrictions across Meta, Google, LinkedIn, TikTok, and display.
VIEW CASE STUDY →
AI & GenAI
Product Leadership
Domain Expertise
Platforms & Tools
AI and Generative AI product leader who can build and ship agents that deliver real user and business outcomes. I translate model behaviour, prompt engineering, guardrails, and agentic architecture into scalable products that users and organisations can trust — closing the gap between what AI can do and what businesses actually need.
Hands-on experience designing and deploying production AI agents including grounded FAQ bots, RAG pipelines, multi-step orchestrators, and multimodal agents. Deep fluency in system prompt design, guardrails, context window management, tool use, and evaluation frameworks. Currently building a portfolio of 18 agents across Retail, FinTech, and developer tooling domains.
Led omnichannel commerce strategy for The North Face, Vans, Timberland, Pottery Barn, and West Elm — owning checkout, payments, loyalty, and personalisation at scale. Delivered $120M+ incremental GMV through conversion optimisation and AI-assisted product discovery. Proven ability to ship in complex, multi-brand, regulated environments.
Wells Fargo — Mobile banking for 8M+ users; fraud detection (−31% losses), bill pay, open banking APIs, PCI compliance.
VF Corporation — Unified Payment Platform (Adyen), GDPR & CCPA, KYC, payment rails modernisation, Klarna BNPL launch.
Metavante Corporation — Foundational fintech: banking tech, payment processing, and enterprise solutions for financial institutions.
Operated at Principal and Head of Product level — defining multi-year roadmaps, building product teams from scratch, and aligning cross-functional stakeholders across engineering, data science, design, and commercial. Experienced in OKR frameworks, SAFe agile delivery, GTM planning, and board-level product storytelling.
Ran 60+ A/B experiments annually across conversion funnels, delivering $40M+ in incremental revenue. Proficient in Amplitude, Tableau, and cohort analysis. Applied LLM-as-Judge evaluation frameworks to measure AI agent accuracy, hallucination rates, and guardrail effectiveness — bringing rigour to AI product quality that most teams lack.
Fluent across the modern AI and commerce stack — Anthropic API, Claude, RAG, vector search, MCP, Cloudflare Workers, Shopify Plus, SFCC, Adyen, Klarna, AWS, MuleSoft. Able to prototype directly, review technical architecture, and partner effectively with engineering on complex integrations without losing speed or commercial focus.