Ipsita Mohanty

• Heading strategy and planning for the WIE SCV chapter • Led the organization committee for the IEEE WIE AI Summit 2025, showcasing leadership and event planning skills.

Building QApp is an end-to-end AI/ML automation platform enabling freelancers, coaches, academies, and consultants to run their businesses seamlessly.

• Led 0-1 GA strategy & execution for AI-driven personalization, influencing cross-functional teams (50+ engineers, researchers, and PMs) to deploy real-time ranking & retrieval NL2VIS Conservational Agentic AI systems resulting with impact of +3XB ARR in revenue and +3XM to customer base expansion. • Drove AI alignment at the executive level, partnering with VPs, product leaders, and engineering and research teams to integrate AI-powered co-pilot and agentic experiences into Salesforce’s enterprise ecosystem. Led hiring & mentorship initiatives, expanding ML teams and improving productivity through structured technical coaching & career development • Managed performance and career development of 5 Staff MLEs for applied science roadmap delivery. • Designed evaluation metrics and generated synthetic data and NLQs for Relevance Scoring and reference-free evaluation using LLMs-as-Judge for NL2VIS. Conversation summarization using In-Context Learning and Fine-Tuned Flan-T5 with PEFT and RLHF (PPO) to generate human preferred summaries. • Architected Agentic Multi-Modal RAG-based QnA for Tableau, designing low-latency retrieval and document-grounded AI insights, improving response accuracy by 15%. (Patent approved for filing.) • Optimized large-scale LLM inference, reducing compute costs by 40% while improving model efficiency for search, ranking, and summarization. Defined experimentation & evaluation metrics, implementing synthetic data pipelines and LLM-as-Judge frameworks to ensure high-relevance content generation.

• Led team of 15 MLEs to develop machine learning and deep learning models at scale, to help accurately forecast revenue and demand for Private Brands business to help merchants effectively quantify buy, decide store assortments, and evaluate category performances • Designed and implemented recommendation and staged ranking systems using 2- tower Deep Learning architecture. • Implemented Deep Learning-based NLP models using state-of-the-art-based embeddings e.g., BERT, RoBERTa, DistilBERT, etc on GPUs for automatic routing of glass project - product reviews defects to concerned engineering teams. • Drove client engagements to identify business problems and deliver business insights using machine learning to help merchant decision-making.

• Modelled Sequence-to-Sequence LSTMs architectures, XGBoost for time series forecasting, and focused on feature engineering to fine-tune the models for accuracy and performance. • Implemented Unsupervised Image Processing Algorithms to extract colors from images. • Implemented KMeans Clustering with PCA for dimensionality reduction for customer and market segmentation. • Designed experiments to evaluate the model performances and identify customer decision makings patterns to improve customer penetration and effectively drive revenue by increasing the sales of products.

• Designed and developed scalable solutions for feeds gateway to catalog item setup. Architected the feeds flows channel for product catalog set up to handle large volume of data ingestion through asynchronous flow handling. • Designed Natural Language Processing capabilities for Internal User Productivity and Knowledge Sharing QnA Chat-bot. • Enhanced features and developed Deep Health Check for critical gateway applications. Technologies - Java, Spring, Cassandra, Kafka, Solr

• Evaluate and research startups for seed investments • Source and interview startup founders for cap table

• Evaluate and research startups for seed investments

• Managed a team of 10 MLEs to successfully extend the Client Analytics Platform, Prime Services to Synthetics desk of Equities, re-vamping the business logic and calculations of Commissions and PnL in the Year-to-Date and Projected Views. • Designed and developed Pricing Engine for performing “what-if” scenario analysis for portfolios of hedge fund clients, helping the sales and management make pre-trade investment strategies and evaluate client onboarding. • Drove client engagement and manage the product development from London location for Fund Analytics- performance analysis against peers & benchmarks for hedge funds clients. • Delivered Prime Services Client-at-Glance, a web application which served as a snapshot of client books and their historical and predictive performance for senior management, sales and finance teams.

• Served as a key developer and business analyst for onboarding Centrally Cleared IRS/CDS/CDX onto GSAM Order Management System. Steered the project deliverables with 50% more efficiency, hence achieving time to market in 10 months duration. • Served as a key developer for GSAM Sentry – Guidelines and Compliance Management, a system for pre-&-post trades compliance checking, during the migration of vendor product Fidessa to an in-house product developed in SecDB.

• Served as a business analyst for Indian telecom regulatory requirement of Multi Number Portability for Videocon

• Served as a key VC++ developer for trading financial application for Deutsche Bank, Singapore • Served as an offshore SAP functional consultant at National Grid, UK during the maintenance phase. Handling daily business activities with customers and providing new feature rollouts and bug-fixes in ABAP.