Amiya James Taggart

Data Analyst & Full-stack Software Engineer

Currently: Building AI compliance tools

About Me

I am a results-driven Analyst and Engineer specialising in building automated data solutions. With a First Class Honours degree in Mathematics from Imperial College London, I combine rigorous quantitative skills with full-stack development expertise in Python, Flask, and modern AI models like Gemini and Claude. I excel at transforming complex manual workflows into efficient, scalable systems that deliver measurable value and significant cost savings for clients in finance, law, and corporate sectors.

My approach combines mathematical rigour with practical engineering solutions. I specialise in transforming time-intensive manual processes into intelligent automated systems, typically reducing processing time by 80-95% while maintaining or improving accuracy. Whether working with investment firms on ESG compliance or legal teams on regulatory analysis, I focus on delivering solutions that scale efficiently and provide immediate, measurable ROI.

How I Work

1

Analyse & Understand

Deep dive into existing workflows to identify bottlenecks and automation opportunities

2

Build & Iterate

Develop modular solutions with continuous testing and stakeholder feedback

3

Deploy & Scale

Implement robust systems with monitoring and documentation for long-term value

Featured Projects

Modern Slavery Benchmark Automation

Live

CCLA Investment Management

Developed a Gemini-powered scoring system that achieved 95% accuracy against manual assessments, reducing a half-year benchmarking process to just weeks per project.

Python OpenPyXL Gemini AI
Saved 6+ months annually

Claims Assessment Web Tool

In Development

Corporates

Created a web application that analyses corporate documents for unsubstantiated ESG claims, generating risk assessment reports with actionable mitigation steps.

Flask Celery RabbitMQ
Real-time analysis

Maritime Regulation Compliance Benchmark

Completed

Law Firms

Built an automated Python system to analyse c.500 regulations, extracting and structuring over 5,000 key compliance insights for clients within two weeks.

Python Document Processing Claude AI
500+ regulations processed

Net Zero Regulatory Analysis System

Completed

UN PRI

Implemented an LLM-based classification system to analyse over 1,000 climate policy instruments, reducing processing costs by 80% through advanced caching algorithms.

Python Caching Multi-threading
80% cost reduction

Core Skills

Programming & Development

Python Flask Celery Pandas & NumPy Git RabbitMQ

AI / ML

Prompt Engineering Google Gemini Anthropic Claude Token Optimisation

Data Engineering & Infrastructure

Data Caching Systems Multi-threading API Integration Document Processing

Technical Showcase


# Example: Efficient document processing pipeline
def process_documents(docs: List[Document]) -> AnalysisResult:
    """
    Demonstrates multi-threaded processing with intelligent caching.
    Reduces API calls by 80% through smart deduplication.
    """
    with ThreadPoolExecutor(max_workers=8) as executor:
        # Process documents in parallel with caching
        futures = [executor.submit(analyse_with_cache, doc) for doc in docs]
        results = [f.result() for f in concurrent.futures.as_completed(futures)]
    
    return aggregate_results(results)
                

This pattern achieves 8x speedup for document processing while maintaining thread safety and reducing API costs.

Career & Education

Dec 2023 - Current

Analyst

Canbury Insights, London

Nov 2023

Professional Certifications

  • Financial Engineering and Risk Management - Columbia University
  • Financial Markets (with Honours) - Yale University

Oct 2019 - Oct 2023

MSci Mathematics, First Class Honours

Imperial College London