A timeline of systems shipped, not just titles held.

Designing a RAG-based compliance intelligence assistant that gives an enterprise product team conversational access to US regulations and safety standards.

• Designing a retrieval-augmented compliance assistant that lets product team members conversationally query new US regulations and safety laws, reducing manual regulatory research.
• Architecting semantic vector search over compliance knowledge bases to retrieve relevant regulatory clauses and map them to product-label requirements.
• Scoping the document ingestion and chunking pipeline that keeps the knowledge base current as new compliance standards are introduced.
• Integrating the LLM-powered chat experience with existing legacy systems used by the product team.

ImpactCuts manual regulatory research for product and compliance teams by turning hundreds of pages of standards into a queryable knowledge base.

• Python
• LangChain
• RAG
• Vector Databases
• Embedding Models
• LLM APIs
• FastAPI
• Docker
• Legacy System Integration

Built machine-learning platforms and agentic automation systems for enterprise clients, from no-code ML tooling to end-to-end AI sales agents.

• Built a no-code dynamic ML framework automating regression, time-series forecasting and anomaly detection — 87% accuracy on dynamic business datasets and a 60%+ reduction in time-to-insight.
• Developed an AI-powered sales agent that automates LinkedIn lead scraping, cold emailing, LLM-based reply handling and meeting scheduling for end-to-end outreach automation.
• Architected a multi-agent framework with CrewAI for travel planning, accommodation management and activity coordination, integrating multiple third-party APIs.
• Delivered containerised, production-grade AI services for enterprise clients.

ImpactReduced manual ML effort for non-technical teams and automated the full outbound sales funnel for sales teams.

• Python
• LangGraph
• CrewAI
• Scikit-learn
• XGBoost
• TensorFlow
• FastAPI
• Docker
• Gmail API
• Google Calendar API
• DeepSeek

Applied computer vision and OCR to logistics workflows, building detection and text-extraction pipelines for parcel processing.

• Developed a shipping-label detection system with YOLOv8, reaching 80% detection accuracy and improving parcel processing speed.
• Integrated PaddleOCR for automated text extraction and product-to-company mapping via GC numbers, enabling streamlined parcel tracking.

ImpactFaster parcel processing and automated tracking through detection + OCR.

• Python
• YOLOv8
• PaddleOCR
• PyTorch
• Computer Vision