Hi, I’m Vinod M 👋 Data Analyst

Problem: Businesses lack real-time financial insights. Solution: Built an interactive Streamlit dashboard with Python and SQL-driven ETL to surface revenue, profit, and trend analytics. Impact: Enabled faster financial reporting and clearer trend-based decisions for stakeholders.

Problem: Product teams need consolidated KPIs for computer vision experiments. Solution: Built a Streamlit dashboard to visualize model metrics, dataset distributions, and deployment logs. Impact: Helped product and engineering teams track experiment performance and prioritize improvements.

Built a public Streamlit analytics dashboard to monitor GRS metrics, enabling quick stakeholder access to key operational KPIs. Featured as a focused analytics project.

Problem: Need a clear professional presence to attract analytics roles. Solution: Built a focused portfolio showcasing analytics, dashboards, and projects with clear business context. Impact: Improved recruiter clarity and centralized contact/resume access.

Problem: Prototyping and experimentation lacked a central showcase. Solution: Built Sushvin AI Labs as a deployed playground for applied projects and demos. Impact: Centralized experimental demos for stakeholder review and demos.

Problem: Local grocers need an online storefront. Solution: Built a deployed e-commerce grocery web app with product listings and cart flows. Impact: Enabled online orders and improved local customer reach.

A Vercel-deployed platform for the Kamadenu Goushala project with public source code.

Problem: Hiring teams require scalable candidate screening. Solution: Built a Streamlit chatbot to automate candidate interactions and basic screening flows. Impact: Reduced initial screening time and improved candidate engagement.

Problem: Businesses need automated customer interactions on WhatsApp. Solution: Built an automation system using WhatsApp APIs and FastAPI to handle support flows, bookings, reminders, and handovers. Impact: Streamlined customer workflows and reduced manual response workload.

Problem: Extracting structured data from unstructured documents is time-consuming. Solution: Built a transformer-based NLP pipeline to extract key-value pairs and normalize outputs for downstream analytics. Impact: Reduced manual data entry and improved extraction accuracy for prototyped datasets.

Problem: Large images slow web workflows. Solution: Built a Streamlit-based image compressor for fast compression and download. Impact: Reduced image sizes for faster uploads and lower bandwidth consumption.

Problem: Static images limit engagement. Solution: Built tools to animate images and export lightweight animations. Impact: Improved visual content quality for demos and portfolios.

Problem: Manual fruit quality checks are slow and inconsistent. Solution: Developed a computer vision model to classify fruit quality using OpenCV and custom datasets. Impact: Automated quality spotting and reduced manual inspection time in prototype environments.

Problem: Traffic monitoring lacked accurate speed estimations. Solution: Developed a CV pipeline to detect, track, and estimate object speed from video streams using OpenCV and tracking algorithms. Impact: Produced reproducible speed estimates used in the published research and prototype demos.