Akash Sharma Full-Stack Data Scientist

Akash Sharma

Full-Stack Data Scientist | ML Engineer

I build production-grade ML systems that transform raw complexity into actionable intelligence.

About Me

With 3 years of experience as a Full-Stack Data Scientist with domain depth in applied machine learning, forecasting, and system intelligence. I design and deploy production-grade models for classification, regression, clustering, and attribution across varied business pipelines, transforming raw complexity into actionable intelligence.

Experienced in building modular, reusable ML frameworks, I handle everything from feature engineering and automated training to model evaluation and explainability using SHAP. I thrive on delivering end-to-end solutions that are not only powerful but also interpretable and efficient.

Skills & Stack

Languages

Python SQL Shell

Databases & Platforms

ClickHouse BigQuery Google Cloud Storage

ML & Data Science

Scikit-learn XGBoost TensorFlow Keras SHAP Optuna Pandas Jupyter

Visualization

Matplotlib Seaborn

Cloud & MLOps

GCP Vertex AI Airflow PySpark

Dev & Orchestration

GitHub CI/CD Bash Scripts

Core Projects / Contributions

Timeline Experience

AI Engineer, RevSure AI

Dec 2024 - Present

  • Full-time role building scalable ML solutions across campaign optimization and marketing analytics.
  • Development of a Marketing Mix Modeling system to optimize ad spend with interpretable response curves.
  • Created a campaign lift analysis module with automated dashboards and SHAP insights.

Data Scientist, ADA-ASIA

Apr 2022 - Dec 2024

Worked across accounts, delivering full-stack ML solutions for diverse business needs.

Account: RevSure AI

  • Developed and optimized machine learning models for lead scoring, revenue projection, and categorization.
  • Created custom transformers for scalable feature engineering and integrated Optuna for hyperparameter tuning, significantly improving model performance (MAPE, F1-Score).
  • Built a modular regressor-classifier framework, accelerating future model development and reuse.
  • Leveraged SHAP for model explainability to ensure stakeholder transparency.

Account: Leading E-Commerce Aggregator

  • Built a predictive model to forecast booking cancellations using XGBoost and baseline logistic classifiers.
  • Conducted thorough EDA and feature engineering (outlier handling, normalization) to enhance accuracy.
  • Enabled actionable insights that minimized revenue loss and improved operations.

Education

PG Program in Data Analytics

Imarticus Learning (2022)

BSc in Information Technology

Pragati College of Arts, Science & Commerce (2020)

Certifications

IBM Data Science Professional Certificate - Coursera

Advanced Python - Squad Infotech

Other Projects

Interests

Music & Focus

Listening to non-lyrical music while working or lifting weights to stay in the zone.

Community & Learning

Hanging out in Python Discord communities to help others, learn, and stay current.

Reverse Engineering

Exploring how games and systems work under the hood, with a fascination for memory reading.

Creative Tooling

Building small, practical tools from scratch to solve unique problems and test ideas.

Logic & Creativity

Combining logic and creativity to script useful solutions and experiment with new concepts.