Predictive Dashboards, Apps & Software Tools

I build specialized software tools that translate complex psychological and linguistic science into accessible, user-centric applications.

Predictive Analytics Portfolio

A unified modeling environment bridging observational data and active clinical interventions.

We build specialized predictive architectures that transform your raw, messy data into actionable tools. Whether you need a secure pipeline to track active patient trajectories in a clinical setting, or an interactive dashboard to model how your product drives outcomes, we provide the underlying rigorous statistics to make sense of it all. The platform consists of two distinct modules tailored for different purposes:

  • Predictive Outcomes Suite: A modular, agnostic prediction software program designed for real-world data. Phase 1, the Predictive Blueprint, securely extracts statistical coefficients from raw historical EMR data using three distinct variable selection methodologies (A Priori modeling, cross-validated Machine Learning, and a Hybrid framework). Phase 2, the Trajectory Tracker, applies these coefficients to score active patients and calculate session-by-session success probabilities. The architecture operates entirely locally for strict HIPAA compliance, accommodating highly flexible outcome variables, custom definitions of clinical success, and a host of additional statistical features without relying on external cloud processing.
  • Empirical Predictive Dashboard: An agnostic and zero-lag interactive app built on advanced longitudinal mixed-effects models. It calculates the impact of an intervention or product use on continuous and binary outcomes. It also allows companies/organizations/labs to input specific user/patient profiles (e.g., a 31-year-old female using a product 3 times a week at a 5mg dose) to predict outcomes when using the product/tmt versus when not. The dashboard features high-quality visualizations of non-linear changes over time and interactive dose-response curves. Can be adapted, eg, to visualize a lab’s own trial or RWE data for grant reporting, marketing, or stakeholder presentations.

PQM: Practice Quality Mindfulness Tracker

Precision measurement for meditation practice.

Most meditation apps just track how many minutes you sit. The PQM app helps you understand the quality of your practice by focusing on two core skills: Perseverance and Receptivity. By also keeping an eye on your mood, you’ll get a better sense of how your sessions might be supporting you in your daily life.

Built directly on the validated Practice Quality-Mindfulness (PQM) scale published in Del Re et al. (2013). It measures Receptivity (how open you are to physical and mental sensations without judging them) and Perseverance (how well you gently return your focus to your anchor when your mind wanders).

Compute.ES: Compute Effect Sizes

Professional Effect Size Calculator.

A privacy-first web app for calculating and visualizing meta-analysis effect sizes (\(d\), \(g\), \(r\), \(OR\)). Features batch CSV processing and interactive distributions.

This tool leverages the compute.es R package to compute the most ubiquitous effect sizes in clinical research. Below are a select few of the derivations utilized within the dashboard:

Effect Size from Means & SDs (MES) \[ d = \frac{M_1 - M_2}{S_{within}} \] \[ S_{within} = \sqrt{\frac{(n_1 - 1)S_1^2 + (n_2 - 1)S_2^2}{n_1 + n_2 - 2}} \] \[ var(d) = \frac{n_1 + n_2}{n_1 n_2} + \frac{d^2}{2(n_1 + n_2)} \]

Effect Size from T-Test (TES) \[ d = t \sqrt{\frac{n_1 + n_2}{n_1 n_2}} \] \[ var(d) = \frac{n_1 + n_2}{n_1 n_2} + \frac{d^2}{2(n_1 + n_2)} \]

Effect Size from P-Value (PES) \[ d = t_{inv}(1 - p/2, df) \sqrt{\frac{n_1 + n_2}{n_1 n_2}} \] \[ df = n_1 + n_2 - 2 \]

Stanford C-Score (Ranking Tool)

Are you in the Top 1% of Scientists?

A live calculator that scrapes Google Scholar to compute your Composite Score (C-Score) based on the Ioannidis/Stanford methodology. Check your global ranking status instantly.

Language & Cultural Immersion

Applications designed to teach authentic language through immersive, science-backed methodology and deep cultural exposure.

Veloz: Language Learning via Language Islands

A scientifically-grounded language acquisition tool.

Veloz uses the “Shadow Loop” method—an audio-lingual technique focused on rapid repetition and phonological loop activation—to help users internalize new languages naturally.

Interactive Slang Explorer

Master real, authentic street Spanish.

A highly visual, interactive guide to the slang, idioms, and street-level expressions they don’t teach you in a textbook. Dive into regional humor, cultural context, and conversational nuances. Test your fluency and begin mastering the authentic Spanish you’ll actually hear on the streets.

The Master Cookbook

Real recipes while learning real Spanish.

An immersive culinary experience that blends authentic recipes with organic language acquisition. Featuring a dynamic hybrid-translation reading mode and an integrated street slang dictionary. Designed for language learners who want practical vocabulary in a real-world context.

R Packages (CRAN)

I have developed and maintain several open-source R packages that serve as foundational tools for the meta-analysis and statistical community.

compute.es

The Gold Standard for Effect Size Calculation

When precision matters, compute.es is the tool researchers trust. It effortlessly converts between dozens of effect size metrics (\(d\), \(g\), \(r\), \(z\), \(OR\)), ensuring your meta-analysis is built on bedrock data.

  • Impact: Cited in 372 studies on Google Scholar.
  • Reach: 273,434 total downloads.

View on CRAN

MAd

Taming Heterogeneity (Mean Differences)

Meta-analyses are messy. MAd brings order to the chaos of dependent effect sizes. It provides robust aggregation methods that allow you to squeeze every ounce of power from multi-outcome studies without violating statistical assumptions.

  • Impact: Cited in 182 studies.
  • Reach: 107,727 total downloads.

View on CRAN

MAc

Correlational Precision

Correlations are not just numbers; they are relationships. MAc specializes in the meta-analysis of correlation coefficients, handling the complex dependencies that standard packages ignore.

  • Impact: Cited in 74 studies.
  • Reach: 51,881 total downloads.

View on CRAN

Contact Request

If your default mail client did not open, you can copy my direct email details below: