Juber Pathan, AI Manager/ AI Developer

Top skills

Additional skills (24)

I’m an AI Team Lead/ AI developer with a proven track record of successfully leading GenAI/ML initiatives at companies like Airbus Group, Ikano Group, ConvaTec, and many others. I’ve delivered solutions addressing different types of subject areas (search, recommendation, optimisation, waste reduction, CRM, marketing and cybersecurity). My current experience involves leading a team of 7 people delivering a retail eCommerce conversational AI solution. I operate in environments where I work with C-suite stakeholders, secure funding, develop LLM/AI driven services, and deliver production-ready solutions that support critical business decisions and value.

hidden
AI Lead/AI Manager
August 2024 - Today (10 months)
Build a Conversational AI solution that is able to showcase correct products from the catalogue and converse with the user
● The solution is able to:
○ Serve product discovery, match/recommend different products
○ Answer queries related to specific products or policies
○ Manage context switching by the user i.e. looking for a jeans and then look for a t-shirt in same session
○ Answer aspirational queries such as “design me a summer outfit”
● Built and implemented Neo4J and semantic search
● Built and implemented context management pipelines which manages products in memory
● Built and an intent driven architecture with an multi-agents framework implementation
● Built and implemented an in memory search service using FAISS
● Built and implemented various prompts supporting various intents
● Oversaw technical progress and development:
○ Recommendation, Prompt caching, hybrid Neo4J and semantic search, LangFuse integration, LangChain agents and Memory management service
○ Maintain code quality and technical implementation of all services
○ Sprint planning, technical roadmap management, proposing new techniques and ideas to enrich the product further, resolution of product requirements and cascade/delegate within the team
● Train junior peers where necessary
FoilAI (Value Retail)
Senior Data Scientist
January 2025 - March 2025 (2 months)
• Developed a XGBoost sales forecasting model per village per brand as part of the initial PoC

• The aim of the PoC was to address the level of inaccuracy and error variance in Value Retail's manual forecasting process
• Incorporated additional datasets and built an automated data ingestion and feature extraction pipeline

• Developed the model at village level with 149 brands and reduced the variance in errors by 50%

• Improved the overall accuracy from 16% to 2% (MAPE)
ConvaTec
Senior Data Scientist
January 2021 - January 2022 (1 year)
• Using ChatGPT to automate the data analysis pipeline via python and create a LLM 'journey' for conducting anomaly detection in an agnostic manner

• Developing and productionising a lead prioritisation model (CLTV based segmentation intelligence) to operate in harmonisation with the internal Microsoft Dynamics CRM platform. The model ranks the incoming leads for the sales team so their follow-up actions could be optimal for generating sales. The model is expected to provide 10% greater effectiveness and 20% greater efficiency
• Conducted analysis on marketing content analysis and tested hypotheses using NLP techniques to gauge churning rate implications of content on patients. Developed a regression model able to predict patient conversion rates
• Developing a hierarchical price forecasting model for ConvaTec and competitors at both geographical district and individual hospital level for the US market. The model is expected to provide better market intelligence and is expected to aid in stopping the average 5% YoY price decline for ConvaTec

Master of Arts in Statistics
MIT
Micro Masters in Statistics and Data Science
First Class in MEng Aerospace Engineering
University of Hertfordshire
First Class in MEng Aerospace Engineering
Diploma of Management
University of Bath
CMI Level 3 Diploma in Management