Data Analyst · Bangalore, India

Network Engineer at Ericsson transitioning into Data Analytics — with real production experience building Python automation pipelines, SQL databases, and KPI reporting systems, backed by a portfolio of 4 end-to-end analytics projects.

Network Engineer — Data & Reporting · Genius Consultants (Client: Ericsson India) · Jan 2024 – Present

Built Python automation that reduced 4 hours of daily manual KPI reporting to 5 minutes (98% reduction). Designed data validation logic for LTE/NR cell health monitoring and persisted results to MySQL for historical tracking and other projects etc.

Python MySQL Power BI SQLAlchemy

End-to-end pipeline merging 6 datasets across 14 countries and 8 industries → 3 clean analytical tables → 9 SQL queries → 5-page interactive Power BI dashboard. Identified 7 high-readiness under-monetised markets as the clearest vendor opportunity in the dataset.

Python PostgreSQL Power BI

Analyzed 3,900 transactions to uncover revenue drivers, customer segmentation, and discount dependency. Young Adults identified as highest revenue segment ($62K); loyalty conversion opportunity quantified.

Google BigQuery Power BI SQL

Analyzed 100,000+ real orders using advanced BigQuery SQL (CTEs, window functions, EXCEPT/INTERSECT). Found 3,779 cities with demand but zero local sellers — a direct logistics optimisation opportunity.

Python MySQL SQLAlchemy openpyxl

Production automation system built at Ericsson for daily LTE/NR zero-traffic cell detection. Applies business rules across KPI datasets, stores results in MySQL, and auto-generates Excel reports for operations teams.

Python MySQL Power BI SQLAlchemy  |  6 datasets · 14 countries · 8 industries · 517 rows

• Global AI adoption doubled from 33.9% → 73.4% between 2019 and 2025, yet only 45% of pilots reach production
• USA investment ($326.8B) is 5× China — North America holds 55.5% of global AI market revenue
• 7 countries (Singapore, Australia, Japan, Canada, South Korea, Germany, France) score 80+ in AI readiness but are significantly under-served commercially — clearest vendor opportunity in the dataset
• Azure is the fastest-growing cloud provider (44.3% avg YoY) — Microsoft's OpenAI investment directly visible as a 2023 inflection point in the revenue data
• India improved AI readiness by +15.8 points in 5 years — fastest improvement of any country tracked

Python PostgreSQL Power BI  |  3,900 transactions · 18 features

• Young Adults (18–35) are the highest revenue segment at $62,143 — marketing budget should concentrate here, not spread evenly
• Business has 3,116 loyal customers but almost no mechanism to convert the 701 returning customers — a direct retention gap
• Hats, Sneakers, and Coats have a ~50% discount rate — nearly half of sales in those categories require a promotion to close, flagging a pricing problem
• Subscribers and non-subscribers spend nearly the same per transaction (~$59) — volume, not conversion, is the revenue driver

Google BigQuery Power BI SQL  |  100,000+ real orders

• 3,779 cities have active buyers but zero local sellers — every order from those locations travels long distance, directly explaining freight cost and delivery time complaints
• 531 cities already have both buyers and sellers at scale — same-day delivery could launch here immediately with no new seller recruitment
• Top 20 customers and top 20 sellers identified for a unified VIP loyalty programme
• Year-over-year order volume grew significantly from 2017 → 2018, with clear seasonal peaks visible in monthly trend analysis

Python MySQL SQLAlchemy openpyxl  |  Built and running in production at Ericsson India

• 4 hours → under 5 minutes — daily manual KPI reporting effort reduced by 98%
• Zero-traffic LTE/NR cells that previously went undetected for days are now caught and escalated same day
• Full audit trail in MySQL enables historical queries impossible with the previous Excel-only process (e.g. "how many times has this cell been zero-traffic this month?")
• Layered business rules (traffic volume + hours + cell state) prevent false positives — operations team gets a clean, reliable escalation list every morning