Nitish C

<p>Assisting in data cleaning, transformation, and KPI reporting for multiple business datasets.<br>
Building Power BI/Tableau dashboards to improve visibility into operational metrics for internal teams.<br>
Supporting reporting automation using SQL, Excel, and Python to reduce manual effort and improve data accuracy.<br>
Collaborating with senior analysts to generate insights and prepare documentation for business stakeholders.</p>

<p>• Generated 25+ weekly reports and 5 executive dashboards using Python and SQL, supporting operational decisions for 200+</p><p>migrant workers.</p><p>• Managed datasets of 10,000+ records via Google Sheets and Freshsales CRM, maintaining 99.5% data accuracy across 15+ data</p><p>fields.</p><p>• Implemented automation solutions for migrant job allocation, reducing processing time by 75% and operational overhead by</p><p>$15K annually.</p><p></p>

<p>Generated 25+ weekly reports and 5 executive dashboards using Python and SQL, supporting operational decisions for 200+ migrant workers.<br>
Managed datasets of 10,000+ records via Google Sheets and Freshsales CRM, maintaining 99.5% data accuracy across 15+ data fields.<br>
Implemented automation solutions for migrant job allocation, reducing processing time by 75% and operational overhead by $15K annually.</p>

<p>Designed a fully interactive Power BI dashboard for Toyota to monitor project timelines, manpower utilization, activity progress, and PIC workload.<br>
Built 15+ visual components, including Gantt-style progress charts, completion trend lines, doughnut status summaries, and KPI cards.<br>
Integrated monthly man-hours, manpower, and activity-level data to provide real-time visibility into project execution across multiple teams.<br>
Improved management decision-making by enabling variance tracking, activity-level bottleneck identification, and PIC-level workload balancing.<br>
Delivered a professional-grade dashboard UI with filters, drilldowns, and branded themes customized for Toyota.</p>

<p>Designed a Google Sheets–based system with 12 Apps Script functions; processed 500+ daily job allocations with a 95% automation rate.<br>
Reduced manual errors by 80% and decreased processing time from 4 hours to 45 minutes through automated workflows and validation.<br>
Developed a centralized dashboard used by 8 team leads to monitor 1,200+ tasks monthly across 3 geographic regions.</p>

<p>Annotated 2,000+ wagon images using Supervisely/Label Studio to build a high-quality segmentation dataset.<br>
Trained YOLOv8-seg and RT-DETR models for wagon damage detection, achieving up to 60% accuracy (mAP50, mAP50-95).<br>
Performed dataset preprocessing, label validation, class balancing, and augmentations to improve model stability.<br>
Conducted training experiments (20–120 epochs) with hyperparameter tuning for learning rate, batch size, and augmentations.<br>
Analyzed validation curves, confusion matrices, and error cases to refine model performance and reduce misclassification.<br>
Documented the end-to-end ML workflow, including dataset preparation, training logs, tuning strategy, and model comparisons.</p>

<p>Analyzed coal price data from 7 global markets over a 24-month period using Python/SQL; processed 50,000+ price points and identified 12 key risk factors.<br>
Built 15+ interactive Power BI/Tableau visualizations tracking monthly procurement spend with 85% forecast accuracy.<br>
Delivered procurement strategy recommendations resulting in a 10% profit increase and 20% coal sales revenue growth within 6 months.</p>

<p>Analyzed 2,600+ steel heats across 8 production lines using Python/Excel; automated quality control for 15 chemical elements with 98% accuracy.<br>
Designed a 4-platform dashboard ecosystem serving 25+ metallurgists with real-time monitoring of 150+ daily production batches.<br>
Established a predictive quality system reducing manual inspection by 60% (saving 120 hours/month) and improving compliance from 85% to 96% across 5 steel grades.</p>

<p>Engineered an 8-page Power BI dashboard tracking 12 KPIs across 3 production shifts for 50+ operators processing 800+ units daily.<br>
Implemented a Python preprocessing pipeline handling 25,000+ records monthly; improved data quality from 72% to 94% through statistical methods.<br>
Identified quality bottlenecks affecting 15% of production: observed 35% higher rejection with 3 specific ink types and achieved significant annual cost reduction through targeted fixes.</p>

<p>Built a Power BI dashboard to analyze material movement and track procured, consumed, and unused quantities.<br>
Cleaned and standardized inventory data by mapping movement types and creating DAX-based utilization metrics.<br>
Identified excess and underutilized materials, providing recommendations that reduced inventory and operational costs.<br>
Delivered insights that enhanced procurement planning and improved stock utilization efficiency.</p>