Implementing end-to-end LLMOps solutions to manage, deploy, monitor, and optimize large language models in production environments, ensuring high availability, scalability, and performance.

Engineered an innovative image detection model tailored for a major US telecommunications client.

Implementing the model to detect and manage inventories of mobile towers, leading to a substantial revenue increase of $.2-.5M annually.

Developed and deployed an LLM-based Conversational Sales Assistant to enable real-time upsell and cross-sell recommendations during customer interactions in a call center.

Spearheaded the implementation of an Agentic AI approach to provide tailored AI Agents recommendations and action plans for sellers using insights from ML model outputs to optimize decision-making and sales strategies.

Spearheaded the development of advanced models for an ed-tech company, implementing AI color contrast and PDF AI auto-zoning using image detection methodologies.

Successfully reduced the workload by 60% through Color Contrast Analysis (CCA) and achieved a 50% reduction in workload with PDF Auto-Zoning, resulting in a total annual saving of $2M.

Built ML algorithms that power millions of users to discover relevant content generating billions of page views.

Created Deep Learning algorithms using NLP techniques to sift through thousands of news articles being published every minute and bring the most relevant content to each individual user.

Spearheaded the development of advanced models for an EdTech company, implementing AI color contrast and PDF AI auto-zoning using Image Detection methodologies.

Developed predictive models using Deep Learning and NLP techniques.

Conceptualized and built an AI-based SaaS product for an innovative AI startup.

Delivered five distinct AI models that empowered marketers with valuable insights, contributing to enhanced marketing strategies.

Pioneered the development of an HR insight tool incorporating advanced logic and Smart JD Procuring using LLM and OpenCV.

Developed a Machine Learning and Deep Learning recommendation engine to production to conditionally recommend similar stories, best stories based on probability for different widgets based on reading history, which tends to increase revenue by 30%.

Implemented feature impact analysis to determine the effect of changing various factors on target variables, resulting in a direct 20% increase in revenue.

Built Artificial Neural Network using TensorFlow in Python to identify the customer’s probability of canceling the connection (Churn Model).

Applied ML algorithms and statistical modeling like decision trees, text analytics, NLP, supervised and unsupervised algorithm, regression models, classification models, social network analysis, neural-net, Deep Learning, and clustering using scikit-learn.

Worked with 8+ country operators, providing ML/DL-based solutions, improving network by 5-40%, and increasing customers by 10-35%.

Worked independently and guided teams on Sentiment Analysis, capacity prediction, throughput prediction, CAPS time series forecasting, etc.

Built a Bert-based review model, sentiment analysis on conversation, and review snippets like advanced 3 NLP/textdata Deep Learning models.

Performed daily analytics activity by monitoring daily KPIs using Citrix and BO for 3 countries.

Worked as an UMTS engineer on a 3G project for Vodafone (rollout and optimization) for Delhi circle.

Did EDA and RCA to get insights from data and worked on KPI optimization and parameter tuning based on statistical models.