Sep 2025 – Present · 9 mos

Sep 2021 – Aug 2025 · 3 yrs 11 mos

• @stackshare:abhra/viva-insights
• 📊 Embedded Power BI Visuals - Architected and Implemented an end-to-end foundation for embedding Power BI reports using the Fabric REST API and JS SDK. Unlocking Power BI embedded visuals obviated the engineering cost of JavaScript-based custom visuals, making us the first Microsoft product to scale Embedded Power BI to 15 distinct product scenarios within 3 months. Extended the platform to support out-of-box Instant reports and Customizable reports within the next 12 months across 22 Azure regions in 12 international locales.
• 💡 Viva Insights Connector & Analysis Services - Consulted on the migration of Insights data store from Azure Analysis Services (PaaS) to Fabric Semantic Models (SaaS), reducing compute and storage costs by 50% while maintaining availability and reliability above 99.95%. Led the engineering effort to introduce 15 new advanced aggregation capabilities in the Power BI connector, accelerating the development of out-of-box templates.
• 💫 Infrastructure Remodeling - Transformed 12+ distributed workflows from poll-based to event-based architecture, reducing latency and compute costs. Designed automated lifecycle workflows to purge stale, invalid visualizations, according to regulatory compliance requirements, reducing storage costs. Implemented a targeted observability approach with custom metrics and alerting to detect anomalies in provisioning.
• 💠 Cross-referencing Organizational Data - Architected the foundational capability of users to cross-reference Viva Insights visuals with external organizational business outcomes.

Mar 2020 – Aug 2021 · 1 yr 5 mos

• @stackshare:abhra/viva-insights
• 📊 Advanced Analytics and Teams Integration - Designed custom analytics accelerators for performing advanced behavioral analytics on Microsoft 365 data - including Productivity Insights, Organizational Network Analysis, Business Process Explorer, Workspace Planning, Customer Relationship Intelligence, and other early-stage technical incubations. Developed integrations of analytics views and intelligent agents with Microsoft Teams Platform, increasing the integration of analytics directly in employees’ daily workflows.
• 🔮 Integration Standards Development - Pioneered comprehensive standards and best practices for API design, scenario testing, and Azure enterprise integration. Collaborated with domain experts to scope and validate new solutions, proofs of concept, and technical action plans, driving significant product improvements.
• 🌏 Reports Internationalization - Engineered a Power BI template internationalization SDK supporting 12 international languages based on Tabular Object Model and OpenXML, drastically improving accessibility and utility of analytics insights for international enterprise customers.

Sep 2019 – Feb 2020 · 5 mos

• 🟢 Runtime State Inspection - Developed administrative inspection scripts for Windows Debugger to expose critical runtime state information, and granular administration of file handles and operations, while maintaining Azure Files service stability. Improve inner loop efficiency by implementing auto-mount of builds to test Azure Files on an Hyper-V images.
• ☢ Incident Triage - Implemented distributed tracing and profiling capabilities for comprehensive monitoring of file operations. Served as On-call Engineer for prompt triage and resolution of enterprise integration bugs.

Jan 2018 – Sep 2019 · 1 yr 8 mos

• @stackshare:abhra/azure-industry
• 📊 Unified Analytics Platform - Designed the foundational industry-Agnostic capabilities of Data Integration, Configuration, Security, Provenance, Monitoring and Telemetry using Azure Core services and Microsoft Power Platform. Extended the foundations with industry-specific accelerators and data transformations using Common Data Model.
• 🧑‍⚕️ Healthcare Data Platform - Engineered a Massively Scalable Data Analytics Pipeline facilitating Structured Streaming, Orchestration Workflows, Rich Interactive Querying & Clinical Reasoning using Azure Data Platform and Dynamics Power Platform. The platform scaled to 50M+ structured Healthcare Data records across heterogeneous formats representing 100K+ patient lives.
• 🔒 Data Loss Prevention SDK - Spearheaded the development of a HIPAA Safe Harbor Compliant SDK for Electronic Health Records in Azure Cloud and Edge. It supported FHIR, HL7, DICOM and Unstructured Text formats (using Azure Cognitive Services).
• 🤖 AI Assisted Diagnosis - Pioneered the transformation of Microsoft AI Network for Healthcare solutions through an AI assistant for Early Detection and Remote Screening of Avoidable Blindness. Designed an SDK to evaluate Image Acquisition Quality and detect abnormalities using Medical Image analysis operations, such as Multiple Object Detection and Semantic Segmentation using Deep Convolutional Neural Networks. The AI assistant was widely integrated into Retinal Imaging Equipment in healthcare institutions.
• 📕 Entity Extraction Engine - Prototyped a Text Enhancement Engine for Extraction of Medical Concepts from existing Clinical Narratives facilitating Clinical Decision Support, Personalized Medicine and Population Analytics.

Jun 2017 – Dec 2017 · 6 mos

• @stackshare:abhra/bing-entity-experiences
• 🔂 Knowledge Graph Node Prioritization - Reduced the compute footprint of the graph processing pipeline by 75%, by using a transitive reduction strategy focusing on most significant entity relationships through graph-based priors and attribute inference. Conducted job graph profiling and annotated query plans in Azure Data Lake to identify cost optimization opportunities, while maintaining completeness of critical functionality and integrity of entity relationships.
• 🔬 Unified Publishing Services Architecture - Unified mainstream and preview graph processing pipelines through a single REST endpoint in Satori Enrichment and Inference Experimentation Portal, accelerating the development agility of data acquisition partners by 60% and reducing integration complexity.
• 🕹 Core Inference SDK - Engineered declarative reasoners and configuration points to enable a modular, extensible approach to graph inference, reducing the need for custom code integration complexity, and making the platform future-adaptable.

Sep 2016 – May 2017 · 8 mos

• @stackshare:abhra/bing-entity-experiences
• 💫 Trending Entities - Boosted freshness of category-wise trending entities in Satori Knowledge Fabric by 400%, using trend detection algorithms based on Wikipedia page reputation and browser engagement, significantly accelerating real-time content relevance
• 📣 Centralized Monitoring Hub - Unified disparate system signals depicting agility, performance and availability into a centralized monitoring dashboard. Orchestrated reliable and rich alerting, automated escalations, granular incident enrichment, providing increased visibility and responsiveness into critical stages through the continuous delivery cycle.
• 📚 Content Parity - Spearheaded a Content Parity evaluation SDK to efficiently execute continuous validations on ever-changing entity content, scaling across millions of entities. Reduced large scale data and ontology migration times by 75%, while maintaining content quality and integrity.
• 🌏 Competitive Analysis - Conducted thorough coverage and quality analysis comparing Bing Snapshots experiences to identify content gaps across its international markets. Pioneered the end-to-end enrichment of Snapshot content in Movies, Software, Events and Awards verticals across all Bing surfaces.

Jul 2014 – Aug 2016 · 2 yrs 1 mo

• @stackshare:abhra/bing-entity-experiences
• 🛰 Entity Experiences Infrastructure - Pioneered the architecture behind release automation, deployment, monitoring and incident response for Entity experiences across 23 international locales, released in Bing.com, Windows 8 Search Charm, Windows 10 Universal Search Bar, Cortana and Bing Mobile
• 🛠 Snapshot Builder SDK - Engineered 50+ declarative graph transformers for content rendering in the Entity Snapshots SDK, used by partners to publish UX-ready content, supporting the verticals for Movies, Sports, Events, and Health on all Bing surfaces worldwide. Revitalized experiences through analysis of data availability, coverage metrics, and data validation processes.
• 🛃 Question Answering - Amplified Bing Answers international coverage by 800% through identification, formulation and normalization of queries with question-answering intent, using a generative AI model trained on culture-specific query patterns, combined with localization and spoken language generation for Cortana TTS.
• 🚦 Ranking and Triggering Models - Modernized the entity models with features from web search metadata, click signals, and user engagement, using FastRank Regression and Classification.