Abhilash P.

Services

• ML systems for finance

  Feature engineering, model selection, and evaluation (walk-forward CV, backtests) for trading, risk, and forecasting.

• Data pipelines & MLOps

  ETL/ELT in Python, orchestration, experiment tracking, and deploy to APIs/dashboards with monitoring.

• Dataverse & Power Platform

  Diagnostics, automation, and performance fixes for Dataverse/Power Platform. Pipelines, connections, and ALM.

• Analytics & decision support

  KPI design, cohort/retention, FPs, forecasting dashboards; clear narratives that drive action.

Recent work & results

• Factor model for BSE mid-caps

  Built a multi-factor ML model (fundamental + macro + momentum) with walk-forward validation and SHAP explainability.

  ResultImproved risk-adjusted returns vs. baseline by 240 bps annualized over a 5-year test.

• Dataverse deployment hardening

  Resolved custom-connector rollout issues and automated solution promotions via service principal and ALM.

  ResultZero-rollback deployments; cut release time by ~60% for a multi-env setup.

• Revenue forecasting for SaaS

  Designed a pipeline and forecasting ensemble (classical + ML) with automatic data quality checks.

  ResultMAPE reduced by 18%; alerts on data drift and anomaly spikes.

Notes & posts