Data Science & Advanced Analytics
Building intelligent systems that turn messy data into measurable impact.
Annual Award
MiQ 2025 🏆
Gold Medalist
Rank 1 🥇
I'm a Senior Analyst at MiQ, working at the intersection of Advanced Analytics, Machine Learning, and Generative AI to solve real-world business problems at scale. I specialize in building end-to-end AI systems, from data pipelines and modeling to production-ready applications that directly influence high-stakes, multi-million dollar media decisions.
At MiQ, I've designed and deployed ML pipelines, LLM-powered automation tools, and analytics platforms that bridge the gap between technical innovation and business impact. My work focuses not just on building models, but on creating scalable, interpretable, and value-driven systems.
I was honored with the Annual Gamechanger Award 2025 - MiQ's highest individual recognition, along with two quarterly awards for driving impactful AI and data initiatives.
What sets me apart is a product-first mindset combined with strong engineering depth. Whether it's computer vision systems, NLP-driven audience intelligence, or GenAI workflows, I approach every problem as an opportunity to build intelligent, scalable solutions end-to-end.
I began my journey in AI through research at the Symbiosis Centre for Applied AI (SCAAI) and graduated as a Gold Medalist (Rank 1) in Electronics & Telecommunication Engineering from Symbiosis Institute of Technology.
2+
Years in AI/Data
10+
Advanced Analytics Solutions Built
$5M+
Incremental Revenue Through Custom Solutions
5+
Awards Won in AI and Data Science
Annual Gamechanger Award 2025 — MiQ's highest individual honor, plus 2 quarterly awards
Gold Medalist, Rank 1 — B.Tech ENTC, Symbiosis Institute of Technology
Built AI systems processing 500M+ daily events
AIR 4 nationally in Computer Vision Hackathon
MiQ Digital
MiQ Digital
Symbiosis Centre for Applied AI (SCAAI)
Production systems, research experiments, and hackathon builds — all with measurable outcomes.
AI / LLMs
Media analysts spent 4–6 hours daily manually compiling campaign performance reports from disparate data sources.
Built a RAG-based LLM pipeline using GPT-4 + custom embeddings over campaign data, with a natural language query interface and automated narrative generation.
ML Systems
Existing audience segmentation ran in batch mode with 24-hour lag, causing missed targeting opportunities in programmatic auctions.
Designed a streaming ML pipeline on Databricks with Delta Lake, processing 500M+ daily bid events using gradient boosting models with sub-second inference.
Computer Vision
Ad creative quality was assessed manually, creating bottlenecks and inconsistent scoring across thousands of creatives.
Built a CNN-based visual quality scoring model trained on 100K+ labeled ad creatives, integrated into the creative approval workflow.
MiQ's analytics data was siloed across 10+ tools with no unified intelligence layer for cross-market insights.
Used by 50+ stakeholders globally
$2M+ in optimized media spend
Classifying user intent from browsing signals required expensive third-party data enrichment.
Replaced $300K/year third-party cost
92% classification accuracy
HR teams manually screened hundreds of resumes, leading to bias and inefficiency.
Ranked 4th nationally at CV Hackathon
500 resumes in <30 seconds
2025
Highest individual recognition at MiQ Digital
2025
Outstanding performance recognition
2024
Outstanding performance recognition
2022
Electronics & Telecom, Pune University
2022
Academic Excellence Award
2021
Led 80+ member technical club
2023
All India Rank 4
2024
Internal Innovation Recognition
Sharing what I learn building AI systems in production.
Hard-won lessons on chunking strategies, embedding models, retrieval quality, and why naive RAG fails at scale.
2025
A deep dive into novelty effects, SUTVA violations, and how to run statistically rigorous experiments in ad-tech.
2024
What I learned transitioning from analytics to building production AI systems — and what I wish I'd known earlier.
2024
Open to senior AI/ML roles, research collaborations, and interesting problems worth solving.