Akaran Sivakumar

Data scientist focused on decision-making under uncertainty using behavioral data.

Current Focus:

I frame problems through data, build statistical models, validate rigorously, and communicate insights that drive better decisions.

Approach

Systems Thinking

I approach problems through the lens of interconnected systems, combining cognitive science principles with rigorous data analysis to understand complex behavioral patterns.

Technical Rigor

Every solution begins with experimental validation, statistical grounding, and careful consideration of edge cases before scaling to production systems.

Decision Framework

I prototype quickly, validate with data, and only scale models that survive statistical and behavioral scrutiny.

Core Competencies

Decision-making under uncertainty

Statistical modeling and validation

Behavioral and language data analysis

Problem framing and experimental design

Uncertainty communication

Technical Stack

Machine Learning & Data

Python

scikit-learn

TensorFlow

PyTorch

R

tidyverse

Stan

Bayesian Modeling

SQL

PostgreSQL

BigQuery

Cloud

AWS

GCP

Vercel

Featured Projects

World Cup Sentiment Analysis

Problem:Understanding real-time public sentiment to inform marketing and engagement decisions during global events

Method:NLP pipeline processing millions of tweets with sentiment analysis

Outcome:Enabled data-driven fan engagement strategies and cultural insights for event organizers

View case study →

Personality & Language Adoption (GABM)

Problem:How personality traits influence decision-making in language evolution and adoption

Method:GPT-4o-mini agent-based simulations with Bayesian analysis

Outcome:2,000+ simulations providing insights for behavioral interventions and language learning strategies

View case study →

Peerly

Problem:Streamlining academic peer review workflows to improve decision quality in publishing

Method:Full-stack platform with automated reviewer matching

Outcome:Real-time collaboration platform reducing review time and improving matching accuracy

View case study →

Experience

Co-Founder

2024 - Current • BetterRoomie, Aarhus

Built a cognitive-science–driven matching platform; designed compatibility algorithms and data pipelines. Funded by The Kitchen, Grundfos, and Fonden for Entreprenørskab.

Data & Marketing Analyst

2022 - Sep 2025 • Curry Leaves, Sønderborg

Improved SEO by 63% and automated reporting processes using Python and Zapier.

Marketing Assistant

Oct 2025 - Dec 2025 • TwelveSixteen, Aarhus

Led Google Ads campaigns and GA4 tracking; improved ROAS through data-driven audience and funnel analysis.

View Full Experience →

Have a project in mind?

Open to data science and ML roles.

Contact Me