Abhijeet Kumar

Results-driven Data Analyst & Analytics Engineer with 2+ years of experience building end-to-end data pipelines, SQL-heavy ETL workflows, and executive BI dashboards. Deep expertise in Trino/Presto, Databricks, dbt, Power BI, and Python-based automation. Proven track record of processing 40M+ lead records, reducing reporting latency by 30–70%, and directly influencing revenue and conversion KPIs.

Skilled at translating complex data into actionable business intelligence for EdTech, sales, and marketing domains.

Data Analyst at Physics Wallah (PW) (2024-12 – Present)

• Architected and maintained gold-layer dbt pipelines for 40M+ lead records (16 GB dataset) using Trino/Presto on Databricks, applying advanced SQL patterns (CTEs, LEAD() window functions, UTC-to-IST conversion) to power realtime counselling performance analytics.
• Designed and executed A/B testing frameworks across counselling outreach strategies and campaign segments, generating statistically significant insights that drove a 20% surge in lead conversions and doubled funnel efficiency.
• Developed target achievement prediction models using historical conversion trends and cohort-level behavioral data, enabling proactive performance interventions and accurate monthly target forecasting for leadership.
• Applied predictive analytics and time-series forecasting on lead flow data to detect seasonal demand patterns and anomalies, reducing reactive decision-making and supporting forward-looking capacity planning.
• Designed and automated Python + MySQL ETL pipelines integrating 5M+ records from APIs and external systems into a centralized data Lakehouse; implemented incremental data loads and schema evolution handling, reducing reporting latency by 30% and boosting data accuracy by 20%.
• Built advanced Power BI dashboards with embedded predictive trendlines and cohort-level drill-throughs monitoring CPL, RPL, Conversion Efficiency, Lead Aging, and agent-level KPIs — cutting reporting cycle time by 70% and driving a 20% productivity increase.
• Engineered statistical anomaly detection pipelines on 40M+ lead records using SQL window functions and Python (NumPy, Pandas), identifying irregular lead flow patterns and behavioral deviations that reduced data noise by 95% and improved funnel integrity.
• Launched WhatsApp + Email KPI automation using Python, delivering real-time performance alerts and predictive target-vs-achievement summaries to leadership — improving reporting transparency by 70%.
• Built Python-based incentive management framework automating performance-linked payout logic with configurable scoring rules, driving 25% productivity gain and eliminating manual payout errors.
• Reduced manual reporting effort by 75% through end-to-end pipeline automation, structured data models from raw event-level data, and reusable dbt model layers.

Data Analyst at Didatic Nest (2022-04 – 2023-05)

• Enhanced inventory tracking efficiency by 40% by streamlining Excel-based CRM and inventory dashboards integrated with Zoho and QuickBooks.
• Built KPI trackers for stock aging and reorder points, reducing stockouts by 25% and lowering storage costs by 10%.

Bachelor of Arts in English Honours – Patliputra University (2018-07 – 2021-12)