Hi, I'm Maheshwor Tiwari
Data Analyst | AI Engineer
Crafting AI Agents & Data Solutions: Turning Complex Analysis into Intelligent Action
About Me
AI Engineer and Data Analyst with a Master's in Data Science and experience building intelligent systems and transforming data into actionable insights. I specialize in Python programming, machine learning, SQL, and data visualization with Tableau and Power BI.
My expertise includes developing AI applications and data solutions—from building AI-powered tools like the CV Agent to creating interactive dashboards and reports. My background in advanced research and data analysis enables me to solve complex challenges and communicate findings through clear visualizations that support business decision-making.
Experience
Professional Experience
- Evaluated and improved AI model responses by assessing accuracy, quality, and alignment with user expectations
- Developed and tested training prompts to identify AI performance gaps and enhance model capabilities
- Provided detailed feedback and evaluation on AI training projects, contributing to the development of more reliable AI systems
Data Analyst & AI Engineer
The Bitrates (Contractual)
Hurstville, NSW, Australia Jul 2024 - Present (Current)
- Analyzed client website databases using PostgreSQL and Power BI, delivering actionable insights that improved forecasting accuracy by 25% and enabled data-driven business decisions
- Developed CV Agent, an AI-powered resume builder using Flutter/Dart frontend, Python backend, and multiple LLM providers, serving 500+ users with real-time ATS optimization
- Collaborated with cross-functional teams to identify business opportunities through advanced statistical analysis
- Automated data extraction for population datasets using Python, improving accuracy by 20% and team collaboration
- Built Power BI dashboards that reduced client reporting time by 30%, from 10 hours to 7 hours per week
- Optimized customer support response times through data analysis, improving operational efficiency by 15%
Software Engineer and Analyst
Property Console
Sydney, Australia Jun 2023 - Nov 2023
- Built Python scripts to track key metrics, ensuring 99% data accuracy and improving data integrity and operational efficiency
- Collaborated with the development team to enhance analytics capabilities, using Python to improve data processing speed by 25%
- Leveraged Python for data preprocessing and analysis, producing Tableau dashboards that improved decision-making
Academic Experience
Lecturer and Course Facilitator
CY Cergy Paris University
Cergy-Pontoise, France Sep 2019 - Jun 2022
- Assisted professor by leading physics tutorials for Master's students (8 hours/week per semester) in Condensed Matter Physics, Hamiltonian Systems, and Many-Body Problems
- Designed course materials, assignments, and assessment rubrics, while mentoring students on research projects and fostering critical thinking through problem-solving sessions
- Collaborated with faculty on curriculum development, creating engaging tutorials that enhanced student understanding and promoted teamwork
Research Intern
CY Cergy Paris University
Cergy-Pontoise, France Feb 2018 - Jun 2018
- Developed Python programming proficiency through computational research on magnetic field effects in 2D materials
- Managed and analyzed complex datasets using statistical analysis and data modeling techniques, improving data accuracy and reliability
- Created Power BI dashboards to visualize research findings and participated in weekly reviews, enhancing technical communication and collaboration skills
Physics and Mathematics Teacher
Rastriya Secondary School
Nepal Jan 2015 - Aug 2016
- Managed and mentored 60+ students across Class 11 and 12 Physics and Mathematics, developing strong classroom management and communication skills
- Developed curriculum integrating real-world data applications, analytical thinking, and problem-solving techniques adapted to diverse learning styles
- Created hands-on experiments and projects connecting theoretical concepts with practical applications, building strong student rapport through individualized mentorship
Featured Projects
Selected work showcasing AI/ML, data analysis, and full-stack capabilities
CV Agent - AI-Powered Resume Builder
AI-powered resume builder with real-time ATS optimization, role-aware tailoring, and skills gap highlighting. Integrates multiple LLM providers with prompt orchestration, supports PDF/Doc parsing, and exports recruiter-ready resumes and cover letters
Impact:
- 500+ users served; consistent engagement on role-specific templates
- AI-driven content suggestions with quantifiable achievement phrasing
- Real-time ATS scoring (keywords, section structure, readability)
- Multi-model fallback for reliability; export to PDF/Doc with templating
YOLOv8n Corrosion Detection Optimization
Research and engineering to optimize YOLOv8n for real-time corrosion detection in resource-constrained environments (drones/edge devices) using structured and unstructured pruning, with fine-tuning to preserve accuracy
Impact:
- 11.39% model size reduction via structured (L1-norm) filter pruning
- 39.34% faster inference; improved edge deployability
- 7.43% mAP@50-95 drop post-pruning, mitigated by fine-tuning
- Demonstrated trade-offs: unstructured pruning risked higher accuracy loss at high sparsity
A Comparative Study for Machine Learning Models for Predicting Obesity
Comparative analysis of Logistic Regression, Decision Tree, and Random Forest models to predict obesity levels using lifestyle and demographic attributes. Includes preprocessing (OHE, outlier handling), EDA, and pipeline-based training.
Impact:
- Random Forest: 95.22% accuracy (best-performing model)
- Decision Tree: 94.10% accuracy at max depth 9
- Logistic Regression: 85.01% baseline accuracy
Diagnosing Parkinson's Disease using Voice Sample Data Analysis
Non-invasive PD diagnosis approach using acoustic features from voice samples. The study analyzes speech-derived indicators and frames a methodology toward a practical, reliable diagnostic tool.
Impact:
- Identified key acoustic markers (e.g., MaxPitch, StdDevPitch, Jitter(%), UPDRS)
- Demonstrated feasibility of non-invasive, voice-based PD screening
- Provides groundwork for ML classification and clinical validation
Education
Master of Data Science
Charles Darwin University
Sydney, Australia 2023 - 2024
GPA: 6.35/7
Thesis: Master Thesis (Data Science)
PhD in Physics (Condensed Matter)
CY Cergy Paris University
Cergy-Pontoise, France 2018 - 2022
Thesis: Mean-field theory for quantum spin systems and the magnetocaloric effect
Master in Theoretical Physics
CY Cergy Paris University
Cergy-Pontoise, France 2016 - 2018
Thesis: Effect of magnetic field in square and honeycomb lattice (Internship Report, 2017–2018)
Core Skills
Let's Connect
Open to opportunities in AI Engineering, Data Analysis, Business Analysis, Data Engineering, and Data Science. Available immediately and willing to relocate.
Direct Contact
Prefer to review my resume first?
Preview CV
