Aparna Wani

Responsible for:

• Design and implement comprehensive solution architectures tailored to project requirements.
• Responsible for the technical delivery of generative AI projects.
• Ensured seamless coordination and delivery of multiple initiatives.
• Stakeholder management across different project phases.

Gen AI Use Cases:

• Delivered customized RAG solutions to multiple industry sectors.
• Implemented nl2sql framework tailored for retail and telecom sectors.
• Designed and implemented a feedback loop to capture the feedback and derive actionable insights from collected feedback.
• Developed an AQM solution for performing post-call quality analysis.

• Free-lanced to set up data science discipline
• Contributed to pre-sales activities and created POCs for clients.

Worked with multiple teams including Sales IT, Azure Quota Management and MCS

Developed Time Series forecasting models for key business needs, including:

• Predicting cloud consumption for Microsoft products (Azure, O365, D365) using the HTS framework, with insights visualized in a Power BI dashboard utilized by 1,500+ monthly active users within Microsoft.
• Forecasting monthly incident volumes to support planning and budgeting for seamless customer support.
• Designed and implemented hybrid/ensemble models to enhance prediction accuracy and reliability.

Developed NLP solutions to address diverse business challenges, including:

• Text classification model for evaluating quality issues needing further investigation, utilizing a limited domain-specific data set of 650 records. Incorporated LIME and SHAP for explainability.
• Entity Relationship Extraction: Designed a solution for pharma researchers to identify relationships between entities within a large corpus, aiding drug discovery. Delivered insights via a Power BI dashboard, and utilized Azure Search and SpaCy for implementation.
• Conversational AI for Telecom: Implemented a chatbot using LUIS and QnA Maker, collaborating with stakeholders to fine-tune training data for optimal performance.

Revenue Optimization and Price minimization for pharma companies – worked on IRP (international reference pricing) and LSO (launch sequence optimizer) solutions which optimizes the revenue for pharma companies by suggesting the order in which the new drug should get launched globally before the patent expires. Used Simulated Annealing algorithm for optimization.

Worked on BI solutions using excel Pivots and charts. Created backward-looking reports to aid business review.

Worked as a tester for the statistical solutions developed for pharmaceutical companies to design and analyze clinical trials, and I worked on Excel macros, SAS programs, and automated testing tools.