Akash M

About Akash M

Microsoft Certified Data and Analytics professional with nearly 5 years of experience building scalable analytics solutions using Power BI, SQL, Azure, machine learning algorithms, and Power Automate.Experienced in data modeling, semantic modeling, ETL transformations, Row-Level Security (RLS), and advanced DAX, delivering enterprise BI solutions across multiple projects.I have also worked on AI and machine learning solutions using Python, including STAN (Support Theme Analyzer) — an AI-driven system for clustering and analyzing large volumes of customer support incidents to generate actionable insights.My work spans Business Intelligence, Cloud Analytics, and AI-driven data solutions, helping organizations transform complex data into meaningful insights for decision-making.

Experience

1. Technology Analyst

   Infosys

   Nov 2024 — Present · Bengaluru, IN

   Developed Power BI dashboards and reports for the Microsoft Commerce Risk Customer Engagement team, writing complex SQL and KQL queries to extract and transform data from analytical sources and use them as datasets for reporting and operational insights.• Automated weekly and monthly manual workflows by developing complex Power Automate flows, reducing manual effort and improving operational efficiency.• Led the design and development of STAN (Support Theme Analyzer), an AI driven analytics platform to group and analyse large volumes of customer support incidents and identify recurring issue themes.• Built the initial clustering pipeline using Python, applying UMAP for dimensionality reduction and HDBSCAN for unsupervised clustering to generate seed clusters representing support issue themes.• Developed a hybrid ML-LLM workflow where Azure OpenAI (GPT-4.1) performs daily classification of new incidents into existing clusters, while periodic unsupervised clustering and cluster deduplication identify new themes and update the theme library.• Generated vector embeddings using text-embedding-3-large to represent support incidents and support the initial semantic clustering pipeline with UMAP and HDBSCAN.• Provisioned and managed Azure infrastructure for the STAN platform, including Azure ML Studio workspaces, Azure Data Lake Storage Gen2 containers, and compute resources, and used Azure Data Share to clone incident data from source Kusto clusters into the STAN analytics Kusto cluster.• Developed Power BI dashboards to visualize STAN outputs, enabling stakeholders to track recurring support issues, theme distribution and emerging problem trends.

Education

• VELS University

  Bachelor of Technology, Petroleum engineering

  2016 — 2020

• St. Joseph's Pre-University College

  PUC, Science

  2014 — 2016

• Great Learning

  Postgraduate in Machine Learning & Data Science, Computer Science