Santan Tanuku

Scudy AI – End-to-End Conversational Assistant

• Architected and deployed a production-grade AI chatbot (Scudy AI) on the corporate website to handle customer service, order processing, and supply chain inquiries.
• Enabled full conversational support for checking product availability, tracking orders, initiating refunds, and resolving operational queries.
• Built using OpenAI API, LangChain, and FastAPI; integrated with internal APIs and databases for real-time response generation.
• Delivered measurable improvements in customer satisfaction and reduced support ticket volume through automation.

AI-Powered Part Discovery Engine

• Designed an AI-driven application to extract, summarize, and contextualize technical part data (SKU-level) from web and internal sources.
• Integrated Python backend with OpenAI API to generate accurate, human-readable outputs on part specifications, alternatives, and sourcing options.
• Deployed vector search and document parsing to reduce manual lookup and accelerate part discovery for customers and internal stakeholders.

AI-Based Dynamic Pricing Engine

• Developed a dynamic pricing system leveraging machine learning to adjust margins based on regional demand, product behavior, and customer segmentation.
• Built predictive models using Python and scikit-learn to determine optimal price points in real time.
• Integrated solution into internal sales systems, enabling automated, data-driven pricing decisions and improving profit margin control.

Business Intelligence, Automation & Analytics

• Power BI Reporting: Designed advanced dashboards across Sales, Marketing, Supply Chain, and Finance; implemented DAX-driven KPIs, interactive visuals, and drill-through analysis for real-time business tracking.
• Predictive Forecasting & LTV Models: Developed forecasting models for marketing performance and customer lifetime value (LTV), supporting strategic budget planning and revenue optimization.
• Data Automation: Automated high-frequency reporting workflows using Python and Power BI REST APIs, reducing manual reporting efforts by 60%+ and accelerating delivery of insights.
• SEO & Digital Performance Analytics: Supported digital marketing strategy through keyword research, SEO performance analysis, and paid campaign optimization. Collaborated with external agencies to refine targeting and boost acquisition efficiency.
• Service Data to Strategy: Leveraged aftersales/service datasets to produce insights for procurement, pricing strategy, and cost-cutting initiatives. Delivered actionable recommendations that informed cross-departmental decisions.

• Customer Sentiment Analysis: Developed NLP models in Python to classify sentiment from multi-channel customer feedback (email, chat, reviews). Enabled early detection of dissatisfaction, driving proactive support and reducing churn.
• Predictive Customer Experience Modeling: Built multi-class classification models to segment customer interactions by pain points and behavioral intent, improving personalization and raising satisfaction KPIs.
• Python-Driven Automation: Automated data ingestion, transformation, and model execution pipelines to enable real-time customer insight generation and agile service optimization.
• Voice of Customer (VoC) Pipeline: Implemented IBM Watson Speech-to-Text API to transcribe and analyze call center audio, generating structured data for trend analysis and quality improvement.
• Exploratory Data Analysis (EDA): Conducted in-depth EDA using Pandas, Seaborn, and Matplotlib to surface behavioral patterns and inform feature engineering and campaign strategies.
• AI Model Optimization: Refined and validated 12 production AI models for customer profiling and support routing; improved precision and confidence intervals for real-time applications.

• Dashboard Development & UX Optimization: Designed and deployed strategic Tableau dashboards for Sales, Finance, and IT departments. Implemented visualization best practices to improve readability, driving a 30% increase in adoption and faster decision-making.
• SQL-Based Data Integration: Created optimized SQL views for real-time data extraction and reporting. Collaborated with data engineers to ensure accurate, clean, and high-integrity data pipelines. Reduced data retrieval latency by 40%.
• User Enablement & Training: Conducted hands-on Tableau training sessions for 50+ business users across functions. Created custom learning resources to promote data literacy and enable a 25% increase in self-service analytics.
• Tableau Server Administration: Maintained 99.9% Tableau Server uptime; optimized performance settings and user access policies, reducing dashboard load times by 20% and enhancing overall usability.
• Technical Documentation & Business Translation: Authored clear documentation and translated business requirements into detailed technical specs. Accelerated BI project delivery by 15% through improved alignment between stakeholders and technical teams.

• Transaction Classification Engine: Built a scalable classification system using Python and SQL to tag transactions by brand, channel, type, geography, and vertical. Streamlined analysis workflows, accelerating decision-making across the business.
• Rule Corpus Design: Co-developed a dynamic corpus of classification rules and pattern recognition logic in Python, enabling high-precision categorization across millions of transaction records.
• Regex Query Optimization: Engineered complex SQL queries with regular expressions (Regex) to extract brand identities from raw transaction text. Significantly improved brand detection accuracy while reducing manual tagging time.
• Automated Merchant Identification via Clustering: Led the development of an unsupervised clustering pipeline in Python to identify unknown merchants in unstructured transaction data. Replaced manual tagging with ML-driven pattern matching.
• Operational Efficiency Gains: Automated key workflows for a 25-member Merchant Intelligence team, eliminating manual pattern detection and saving hundreds of hours per quarter.
• Scalable Brand Data Framework: Extended brand intelligence systems to handle millions of merchant records with minimal human input, ensuring consistent and reliable data ingestion and analysis.
• Geolocation Corpus Automation: Built a USA-wide location corpus in Python to normalize geographic references across transaction data. Automated standardization and validation of city/state/ZIP fields for integration into MLC (Master Location Capture) models.
• Custom Spell-Checker for Location Data: Developed a Python-based spell-check algorithm to detect and correct errors in city names, cross-validating with ZIP and state codes to enhance data quality and matching accuracy.

• Statistical Analysis & Insight Generation: Analyzed large and complex datasets using Python and R to uncover trends and deliver actionable business insights that supported strategic decision-making.
• Data Cleaning & Validation: Performed in-depth data preprocessing, ensuring high accuracy and reliability across all datasets prior to analysis. Laid the groundwork for dependable reporting and modeling.
• Data Visualization & Stakeholder Communication: Built interactive dashboards and reports to convey technical insights to non-technical audiences. Enabled faster and more informed decisions across departments.
• Feature Engineering & Model Optimization: Selected high-impact variables for ML models; fine-tuned model performance through parameter tuning and cross-validation, improving predictive accuracy and analytical precision.