AtiarKhan
PIM Engineer and AI Generalist
Product information, governed. Data quality, enforced. AI, only where it removes real work.
SCROLL TO EXPLORE
Built the databases.
Ran the operations.
Now orchestrating the intelligence.
Stored procedures. Store management. AI orchestration. It sounds like three different careers, but each one made the next one better.
Career Arc
2003 – 2011
Where it started
Eight years writing code for real clients — database design, backend systems, data pipelines. Across four companies I learned what it takes to ship software that people actually depend on.
2013 – 2021
Running the business
Led multi-site retail for the best part of a decade (BWS and Dan Murphy's), including a period supervising store managers as relief Area Manager. Full P&L, forecasting, hiring, compliance, teams. Data standards only hold when the people entering the data understand why they exist — I learned that on the floor.
2021 – Present
Owning the product data
Trained into product data from the systems side (SAP, Blue Yonder, PIM), validated records by hand, then defined the rules. Now PIM Engineer at Endeavour Group — owner of the PIM platform behind every product record for two national banners. I know what happens when the data is wrong, so I write the rules that stop it.
For senior and lead roles
Built for the room where product data decisions get made.
[ platform authority ]
Hands-on owner of the enterprise PIM platform for two national banners — Informatica-certified in Product 360 Configuration and Management. Every product record passes through rules I wrote.
[ governance, not firefighting ]
A habit of turning recurring data-quality escalations into IDQ rules, standards and process — fixing the cause, not the record.
[ leads people ]
Eight years leading multi-site retail teams and store managers, then upskilling data teams on PIM. Standards hold when people understand why they exist.
[ ships in public ]
Nine live products outside work — proof of follow-through, not a list of intentions.
Expertise Domains — hover to explore
End-to-end PIM platform configuration — schema design, attribute governance, workflow orchestration across Dan Murphy's and BWS eCommerce. Every taxonomy decision reflects how products actually move through stores.
Enterprise MDM ensuring single source of truth across SAP, PIM, and Traders. Designed reconciliation engines with confidence scoring. Years of seeing bad master data cripple operations drives my zero-tolerance approach.
Complex IDQ rule sets enforcing business logic at ingestion — field-level validation, referential integrity, automated cleansing. I know which errors cause shelf gaps and pricing incidents because I managed the stores.
Parameterised reporting layers transforming raw telemetry into executive insights — eCommerce funnels to inventory velocity. I build dashboards that answer the questions operators actually ask.
Over a decade commanding high-volume retail operations across Dan Murphy's and BWS. This operational fluency is why every PIM rule, data quality check, and AI model I build solves real floor-level problems.
Multi-agent AI with strict execution flows (PoC stage) — CCTV hazard detection, knowledge retrieval bots, autonomous data reconciliation. AI for retail safety isn't abstract — I've walked the floors and managed the incidents.
Eight years building production systems across IMS-Health (IQVIA), Informatica-BD, Probity, AJI-Group — relational schemas, stored procedures, DTS-SSIS migrations, server infrastructure.
Full-cycle web delivery — server-side architecture, API integration, frontend, documentation. Now applying that integration discipline to connect PIM, SAP, GCP, and AI into cohesive ecosystems.
Where each skill was actually used — hover to trace
- [8]Data Reconciliation
- [5]JavaScript / TypeScript
- [5]Python
- [3]Data Governance Frameworks
- [3]GCP BigQuery
- [3]KPI Design and Stakeholder Reporting
- [3]Multi-Agent AI Systems
- [3]NLP / NLP-to-SQL
- [3]Tableau / Looker Studio
- [2]Flask / Web Servers
- [2]Prompt Engineering
- [2]RAG / LLM Integration
- [2]SQL / MS-SQL / MySQL / Oracle
- [1]Adobe Analytics
- [1]Agentic AI (Google ADK, LangGraph, CrewAI)
- [1]Ataccama
- [1]Collibra
- [1]Computer Vision (YOLO)
- [1]DQ Monitoring and Remediation
- [1]Informatica IDQ Rule Design
- [1]Informatica P360 10.5 (on-premises)
- [1]Process Improvement
- PIM Data Quality Rules and Governance[ edg ]
- Agentic AI Safety Management System[ edg ]
- AI-Powered CCTV Hazard Detection[ edg ]
- Confluence Knowledge Retrieval Chatbot[ edg ]
- AI-Enhanced Database Chatbot[ edg ]
- Advanced Analytics Dashboards[ edg ]
- Data and Schema Optimisation[ edg ]
- PriceBloke AU[ live ]
- AI Workshift AU[ live ]
- SafeScroll AU[ live ]
- ASEPS[ live ]
- NeuroReset AU[ live ]
- AusHealthLens[ live ]
- SafeVic[ live ]
- Accord[ live ]
- Commz[ live ]
Counts in brackets are the number of projects each skill was used in. Live products open in a new tab.
The timeline of ongoing knowledge building journey.
Many years across software engineering, retail operations, and data — and each one made the next one sharper.
PIM Engineer — Digital and Data Platform
Endeavour Group (EDG)
Own the enterprise PIM platform (Informatica) behind every product record for Dan Murphy's and BWS — users, roles, organisational structures, attribute and classification models, workflow
Design and maintain IDQ rules that encode product data business logic — converting recurring quality failures into enforced validation rather than repeated manual correction
Escalation point for production data quality issues from centralised PIM users; work back to the source team to fix the cause, not the record
Agree attribute definitions and validation logic with Commercial, Merchandising and Technology stakeholders so standards are accepted before they are enforced
+3 more contributions
Product Data Specialist — PIM
Endeavour Group (Support Functions)
Entered and verified product data in the PIM application — accountable for accuracy and completeness at record level
Defined product data quality rules and thresholds (accuracy, completeness, consistency) in Ataccama and Collibra with the enterprise data governance team
Reconciled SAP, PIM and Traders sources in BigQuery to identify, explain and resolve discrepancies; worked with suppliers to fix issues at source
Streamlined high-volume digital asset management between content systems and customer-facing platforms
+1 more contributions
Space Optimisation Specialist
Endeavour Group (Support Functions)
Engineered data-led planograms in Blue Yonder to optimise shelf-share and category growth for Dan Murphy's and BWS
Translated customer-flow data into store layouts that lifted conversion and shopping efficiency
Product Data Management and Product Lifecycle — Training Pathway
Endeavour Group
Structured transition from operations into product data: SAP master data, Blue Yonder space and range, and the PIM application
Learned the end-to-end product lifecycle — new line setup, attribute standards, range review, publishing readiness — from the systems side after a decade on the floor
Foundation for every governance rule I have written since: standards only hold when the people entering the data understand why they exist
Multi-Site Store Manager and Relief Area (Group) Manager
Endeavour Group (BWS / Dan Murphy's)
Supervised store managers across a group as relief Area Manager; held them to corporate standards and policy
Trained, developed and performance-managed managers; ran succession planning and development conversations
Kept a safe, positive work environment across sites through a volatile period
Multi-Site Store Manager
Endeavour Group (Dan Murphy's and BWS)
Ran high-turnover stores exceeding $150K per week with full P&L, demand forecasting and labour optimisation
Led multi-site teams; recruitment and talent development lifted retention and safety scores
Maintained lean inventory and zero budget variance through periods of high market volatility
Introduced process and compliance improvements and led the change management behind them
Software Engineer (Overseas)
IMS Health (IQVIA), M&H Informatica, Probity Systems, AJI Group
IMS Health: pharmaceutical data platforms for Pfizer, Sanofi, J&J, Novartis, AstraZeneca — MS SQL, SSIS/DTS pipelines, Java
Probity Systems: designed and built the American Teachers Accreditation and Training platform end to end — workflow, rostering, reporting, servers
Database design and query optimisation across MS-SQL, MySQL, Oracle
Full-stack delivery in Java, .NET, PHP, JSP; Apache and SQL Server infrastructure
+1 more contributions
Never stopped learning.
2013 · Central Queensland University
Master of Information Technology (Software Development)
A Master's in IT with a software development focus and path to a Research Paper.
Research Proposal
A Decentralised Multi-Agent Framework for Real-Time Inventory Sustainability in Omnichannel Retail
Autonomous agents coordinating stock across online, in-store and warehouse channels — self-correcting supply chains without human intervention
This is where I am pushing next — can autonomous agents coordinate real-time inventory across every retail channel without human intervention? I know about this Issue in-depth. That is why it is my research.
Professional Certifications
21 credentials spanning product information management, data analysis, systems engineering, cloud architecture, and AI ethics, safety and product management. Still adding to the list.
Systems I have built. Problems I have solved.
Part of a team or solo.
Every project here came from a real problem I ran into at work or in research. I built them because I needed them.
PIM Data Quality Rules and Governance
Reduced manual data-quality escalations and time-to-market on new product listings; single governed source feeding eCommerce, SAP, and analytics
Owns the IDQ rule sets that encode product data business logic for the Dan Murphy's and BWS catalogue — accuracy, completeness and consistency enforced at ingestion, with thresholds first defined in Ataccama and Collibra. Turns recurring quality escalations into rules and process instead of repeated manual fixes.
Reduced manual data-quality escalations and time-to-market on new product listings; single governed source feeding eCommerce, SAP, and analytics
Agentic AI Safety Management System(PoC)
Eliminated hallucinations via strict execution flow
Multi-agent AI system with root orchestrator managing sequential workflows across specialised sub-agents for safety management.
Agent Architecture
Eliminated hallucinations via strict execution flow (store → footage → dataset → policy → report)
AI-Powered CCTV Hazard Detection (PoC) — Winner, EDG AI Multi-Agent Project Competition
Real-time hazard identification from visual inputs
Video analysis pipeline using YOLO for real-time hazard detection from CCTV footage.
Improved detection accuracy by aligning visual signals with domain-specific logic
Confluence Knowledge Retrieval Chatbot(per)
Reduced enterprise information overload
AI chatbot using RAG to retrieve and summarize content from large Confluence datasets.
Reduced information overload and improved enterprise productivity
AI-Enhanced Database Chatbot(Internal Project)
Natural language to structured database insight
Natural language interface to query structured MySQL databases with AI-enhanced readability.
Converted unstructured queries into meaningful, readable insights
Advanced Analytics Dashboards
Executive decision-making at eCommerce scale
Enterprise dashboards for eCommerce analytics and stakeholder reporting.
Enabled data-driven decisions with calculated metrics and KPIs
Data and Schema Optimisation
Overcame row-level limitations at volume
Optimised large product datasets to overcome MySQL row size limitations.
Improved schema scalability and performance for high-volume data systems
Live Deployments — 9 products in public
Accord
Voice meeting intelligence — record, transcribe, decide
One-tap meeting capture that turns spoken conversations into structured outcomes: transcript, summary, decisions, and action items, saved to a searchable archive. Rebuilt as a single FastAPI service with Gemini and Supabase after a multi-service first version proved too complex to run — a deliberate simplification.
Commz
A communication bridge between Deaf and hearing people
Speech-to-text with vibration alerts, text-to-speech replies, quick phrases for everyday, health and emergency situations, a one-tap emergency card, adjustable text size, private conversation history, and offline use once loaded. No sign language required from the hearing side.
PriceBloke AU
Cross-retailer grocery price comparison for Australia
Compares prices across Woolworths, Coles, and Aldi by state. Scan receipts, build shopping lists, track specials.
AI Workshift AU
AI displacement risk scoring for 732 occupations
Scored 732 Australian occupations for AI risk across 13 sectors. Data-driven, research-based, fully transparent methodology.
SafeScroll AU
Independent social media risk scorecard for youth safety
Independent risk tracker scoring major social platforms for child safety. Data from eSafety Commissioner, ABS, and Lancet Digital Health. Built for parents, educators, researchers, and young people.
ASEPS
Australia's first evidence-based state economic growth simulator
Interactive simulator combining ABS and RBA data with Solow-Lucas growth models to diagnose economic gaps across all Australian states and territories. Delivers policy pathways with Victoria at full depth — 10 issues, 24 pathways, and 167 step-by-step workflows. Useful for researchers and students.
NeuroReset AU
A Science-Based Two-Phase Cognitive Reset Model for Australians
Guided stress reduction and cognitive recharge using physiological sigh breathing, 432/528Hz tones, and 40Hz gamma stimulation. Built on peer-reviewed neuroscience from Stanford and MIT. Free, no login, no data collected.
AusHealthLens
Public health intelligence dashboard for Australia
Research and academic tool providing a lens into Australia's public health open-source data. Built from publicly available Australian government datasets for educational purposes.
SafeVic
Victorian crime statistics research platform
Research platform visualising Victorian crime statistics from the Crime Statistics Agency. Explore offence trends, incident data, and regional patterns across Victoria. Built for researchers, policymakers, and the public.
Nobody can show PIM work in public. So I built a model of it.
A fictional catalogue, a real attribute model, and ten data quality rules of the kind I design for product data governance. Toggle a rule. Edit a record. Watch publish-readiness change. Every rule has a reason a store manager would accept.
5/12
[ publish-ready ]
99%
[ completeness ]
10
[ errors ]
5
[ warnings ]
| SKU | Product | Compl. | Status |
|---|---|---|---|
| 104211 | Southern Reach Lager 375ml | 100% | [ publish-ready ] |
| 104212 | Southern Reach Lager Case | 100% | [ publish-ready ] |
| 20871 | Harbour Fog Hazy IPA 440ml | 100% | [ blocked ] |
| 208715 | Harbour Fog Midnight Stout | 100% | [ blocked ] |
| 330019 | Red Gum Ridge Shiraz 750ml | 100% | [ publish-ready ] |
| 330020 | Red Gum Ridge Shiraz Case | 100% | [ blocked ] |
| 441007 | Cloud Line Sauvignon Blanc | 92% | [ blocked ] |
| 552301 | Botanic Yard Dry Gin 700ml | 100% | [ publish-ready ] |
| 552399 | Botanic Yard Dry Gin 700ml | 100% | [ blocked ] |
| 661140 | Glenmorrow 12 Year Single Malt | 100% | [ publish-ready ] |
| 770055 | Southern Reach Zero 375ml | 100% | [ blocked ] |
| 770056 | no name | 92% | [ blocked ] |
[ disclaimer ]
This sandbox is an independent demonstration built for illustrative purposes only. All products, brands, SKUs, GTINs and attribute values are fictitious demo data or derived from publicly available information, and do not represent the data, systems, configuration or business rules of any current or former employer or client. Any resemblance to real products is coincidental. It runs entirely in your browser and stores nothing.
Hiring for a role? Paste the description.
Get a fit score against my actual record — every match cites a real project or role, and the gaps are stated plainly. It runs on the same data Robo-Ati uses. Job descriptions are not stored; up to 10 analyses per visitor per day.
The analysis will show a score, the strongest evidence for each requirement, the honest gaps, and what to ask Atiar about in an interview.
Think you can challenge my expertise?
Select your interview level. Get real questions that probe the depth of the skills showcased in this portfolio. Rate whether the questions challenge you — or if you need a harder tier.
It is a strong commitment to my curiosity.
A Decentralised Multi-Agent Framework for Real-Time Inventory Sustainability in Omnichannel Retail
Autonomous agents coordinating stock across online, in-store and warehouse channels — self-correcting supply chains without human intervention
Real-Time Inventory Agents
Autonomous agents coordinating stock levels across online, in-store, and warehouse channels
Omnichannel Fulfilment Routing
Intelligent order routing that optimises delivery path, cost, and speed in real time
Self-Correcting Supply Chain
Agents that detect stock-outs, reroute inventory, and rebalance without human intervention
This is where I am pushing next — can autonomous agents coordinate real-time inventory across every retail channel without human intervention? I know about this Issue in-depth. That is why it is my research.
If you have read this far, let us have a chat and build something together.
I am always keen for a conversation about data or AI.
Designed and engineered by Atiar Khan
Next.js · React · Three.js · On-device AI · Supabase · Vercel
[ agent-ready · 7 tools at /api/mcp ]
© 2026 — Something for myself