Nishitha Dinesh

Data Scientist at Ashley Furniture Industries (2025-11 – Present)

Architected and build AIDA-Sketch, a hybrid visual search/retrieval system powering Ashley's GenAI based room vision app; sole-owner, shipped phase 0 to internal users with strong reception, subsequent phases now in development for customer rollout.
Engineered dual-search architecture using weighted multi-field scoring and RRF with Azure AI.
Led replacement-parts KPI improvements with quantity-based performance metrics; analyzed historical demand, recommended ratio-based measures, and deployed quantity-driven KPIs that improved forecasting accuracy and inventory decisions.
Implemented graph data science algorithms for route optimization and analyzed customer order and replacement-parts trends to support an anomaly detection use case that flags and helps prevent parts abuse.
Developed a cart-loading tracking solution in Ashley's warehouse using OpenCV to detect AprilTags mounted on carts; updated real-time cart location and status on an edge device by publishing MQTT status messages via Node-RED.
Experimented with multimodal embedding architectures and vision-language models (vLLMs) to expand AIDA-Sketch capabilities beyond visual search/retrieval into sketch generation, improving semantic alignment between query inputs and retrieved assets

Data Science Analyst at Ashley Furniture Industries (2025-01 – 2025-11)

Engineered a regression model in Databricks—leveraging demographic variables from Placer API and historical footfall data stored in the Enterprise Data Warehouse—which reduced forecasting error and guided cross-functional teams on relocating existing stores and siting new ones to maximize traffic.
Utilized Databricks and AutoML to determine optimal parameters for a KMeans clustering algorithm, segmenting regions to recommend new manufacturing site locations; supplemented analysis with a Power BI dashboard visualizing global economic indicators (CPI, GDP, labor force) from FRED to support executive decision-making.
Led supply-chain analytics by extracting order-level data for 460 SKUs over 53 weeks from the Enterprise Data Warehouse, performing demand forecasting, and safety-stock calculations; presented findings through interactive reports and actionable reorder recommendations to supply-chain leaders.
Conducted product-segment analysis for Walmart expansion by integrating regional demographics from Placer with Python scripts and SQL-based aggregation tables to compute performance and opportunity scores; delivered interactive reports to merchandising teams to inform data-driven assortment strategies.

Strategy Data Analyst at Ashley Furniture Industries (2024-06 – 2025-01)

Developed a Power BI dashboard to analyze operations across 100+ retail stores, identifying underperforming locations and improving weekly rankings.
Conducted stocking and backorder analysis for 460 products using Power BI, uncovering weekly trends to optimize inventory levels.
Aggregated and transformed nightly report data to visualize performance metrics for 120 retail stores, enabling real-time monitoring of operational health.
Compiled a comprehensive risk assessment report to forecast budgetary impacts and store traffic for Integrated Business Planning (IBP).

Software Engineering Intern at DRDO (2021-10 – 2021-12)

Contributed to the development of the ABHYAS flight program, reducing the development cycle time by 20%
Integrated IBM DOORS, C++, and LDRA into project workflows, enhancing software reliability and compliance by 15%
Collaborated with interdisciplinary teams to foster knowledge sharing and strengthen cross-functional cooperation.

About