Tushar Y.

Spanda AI provides AI strategies, custom models, and educational solutions embedding intelligence in academia.

• Designed and implemented a data driven AI itinerary recommendation engine using NLP, Sentence-BERT and cosine similarity with Yelp API data on AWS, improved recommendation relevance by 23% and boosted small business discovery in San Bruno.
• Deployed a RAG pipeline with LangChain and LLaMA on GCP, leveraging CI/CD to cut response latency by 40% and enhance chatbot accuracy by 35%, enabling real-time, insight-driven itinerary recommendations.
• Led cross-functional collaboration between developers and business teams to design a data-informed product launch roadmap, orchestrating phase-wise feature rollouts that aligned technical milestones with business objectives.

• Building an AI itinerary recommendation engine combining NLP, cosine similarity, graph algorithms, and Yelp API data stored in Airtable to generate hyper-local, minimal-transit experiences; enhancing and supporting San Bruno’s small business ecosystems.

• Designed and implemented a data driven AI itinerary recommendation engine using NLP, Sentence-BERT and cosine similarity with Yelp API data on AWS, improved recommendation relevance by 23% and boosted small business discovery in San Bruno.
• Deployed a RAG pipeline with LangChain and LLaMA on GCP, leveraging CI/CD to cut response latency by 40% and enhance chatbot accuracy by 35%, enabling real-time, insight-driven itinerary recommendations.
• Led cross-functional collaboration between developers and business teams to design a data-informed product launch roadmap, orchestrating phase-wise feature rollouts that aligned technical milestones with business objectives.

SF Fire Credit Union is dedicated to financial security, serving both firefighters and the wider Bay Area community.

• Led the credit union’s first A/B testing pilot on the member logout page, using Google Analytics 4 funnel analysis to segment traffic and measure conversion lifts, resulting in a 30%+ sustained increase in app downloads.
• Developed a KPI-driven analytics framework for the SF Giants sponsorship event, enabling faster ROI analysis, member acquisition tracking, and cost/reach and cost/lead benchmarking through Salesforce, GA4, Hootsuite, and Metadata integrations.

• Led the credit union’s first A/B testing pilot on the member logout page, resulting in a 30%+ increase in app downloads and improved engagement tracking using Google Analytics 4 to segment traffic, analyze conversion funnels, and optimize CTA placement.
• Developed a KPI-driven analytics framework for the SF Giants sponsorship event, enabling faster ROI analysis, member acquisition tracking, and cost/reach and cost/lead benchmarking through Salesforce, GA4, Hootsuite, and Metadata integrations.

• Optimized targeting and campaign performance for 225+ top-tier clients (Uber, Apple, Coca-Cola, BMW etc.) across 12+ ad categories during IPL 2024, using Excel (pivot tables) & internal ad management tools; achieved a 98% client satisfaction rate.
• Consolidated and validated ad performance data (impressions, engagement metrics) to deliver actionable insights for media agencies, contributing to a 50% increase in repeat business, including for major events like the Paris 2024 Olympics.

JioStar combines Viacom18 and Disney Star, offering extensive TV channels and streaming platforms.

• Optimized targeting and campaign performance for 225+ top-tier clients (Uber, Apple, Coca-Cola, BMW etc.) across 12+ ad categories during sport events, using Excel pivot tables & internal ad management tools; achieved a 98% client satisfaction rate.
• Analyzed and consolidated ad performance metrics (impressions, engagement, conversions) to produce actionable recommendations, driving a 50% increase in repeat business, including for major events like the Paris 2024 Olympics.

• Optimized targeting and campaign performance for 225+ top-tier clients (Uber, Apple, Coca-Cola, BMW etc.) across 12+ ad categories during sport events, using Excel pivot tables & internal ad management tools; achieved a 98% client satisfaction rate.
• Analyzed and consolidated ad performance metrics (impressions, engagement, conversions) to produce actionable recommendations, driving a 50% increase in repeat business, including for major events like the Paris 2024 Olympics.

DK & Associates streamlines regulatory compliance for top-tier manufacturing and real estate clients in India.

• Built environmental compliance dashboards using Power BI, streamlining regulatory reporting processes for India’s top manufacturing and real estate clients; enabled faster audit submissions and enhanced decision-making during production phases.
• Maintained quarterly license records and compliance trackers, avoiding regulatory penalties and contributing to a 100% client retention rate; led cross-functional collaboration between client teams and government regulatory bodies.

• Built environmental compliance dashboards using Power BI, streamlining regulatory reporting processes for India’s top manufacturing and real estate clients; enabled faster audit submissions and enhanced decision-making during production phases.
• Maintained quarterly license records and compliance trackers, avoiding regulatory penalties and contributing to a 100% client retention rate; led cross-functional collaboration between client teams and government regulatory bodies.

Maruti Suzuki is a leading automotive manufacturer in India, known for producing a wide range of vehicles and promoting the 'Joy of Mobility'.

• Conducted root cause analysis on robotic systems using statistical failure analysis, identifying 6 key factors contributing to unplanned downtime, preventing loss of over 1,000 vehicles annually ($5.8 million) and leading to enhanced operational efficiency.
• Developed a Python-based torque anomaly detection tool using pandas and matplotlib, and visualized predictive insights in Power BI, eliminating reducer failures, reducing process time to 30 minutes and saving $10K+ annually.

• Conducted root cause analysis on robotic systems using statistical failure analysis, identifying 6 key factors contributing to unplanned downtime, preventing loss of over 1,000 vehicles annually ($5.8 million) and leading to enhanced operational efficiency.
• Developed a Python-based torque anomaly detection tool using pandas and matplotlib, and visualized predictive insights in Power BI, eliminating reducer failures, reducing process time to 30 minutes and saving $10K+ annually.