we help medical device companies develop predictive models and analytic tools.
Services
What we do
We work with medical device and life science companies to develop and deploy predictive models and analytic tools, in support of diagnostic decision making.
We are a boutique data science consulting company, with deep expertise in biology and statistics. Delv Bio was founded by Jabus Tyerman (learn more about him here).
Machine Learning
We translate measurements and sensor data into scientific insight and provide diagnostic decision support.
We do this by engineering features from your sensor data, and use these features in machine learning to predict ground truth data with high accuracy and precision. These predictions enable diagnostic and decision support adding value to your analytics.
"Jabus is a highly skilled data scientist whom I cannot recommend strongly enough. He has expert knowledge of machine learning approaches, combined with mastery of biology and physiology, making him perfectly suited for the types of research at Respirix. Jabus was able to materially improve our existing predictive models in record time without overfitting, and helped explain our results to internal and external stakeholders."
Eric Kriegstein, CEO, Respirix
"His analysis identified questions and trends we did not notice. His insight allowed us to decide our next steps and efficiently design experiments."
Catherine Pujol-Baxley, VP Research, KnipBio
Engagement
Getting Started
Our engagements always start with a diagnostic data audit that ranges in price from $2,000 - $4,000 (depending on data volume and complexity). This takes 1-2 weeks.
A typical client should expect to spend between $15,000 - $30,000 per project. If this is a good fit for your organization, please contact us by email. We look forward to hearing from you!
Diagnostic Data Audit
Delv's Diagnostic Data Audit is an introductory service designed to quickly assess the problem space and conduct a preliminary data audit.
All our engagements start with this diagnostic because we believe it is our duty to assess the suitability of the data for the problem before recommending a course of action and executing a plan.
Here's how it unfolds:
- Problem space assessment: What are your goals? What are the metrics of success?
- Data Audit: Are the data suitable for the stated goals? What is the status of the data? (organization, complexity, volume, completeness, location). This will include reviewing a data dictionary with the client (all variables in all tables, meaning and units, etc.).
- Recommendations: a summary of our diagnostic and map of next steps.
Next Steps
Every engagement is different, and Delv will use the Diagnostic Data Audit to recommend next steps.
A project may require an exploratory analysis to develop insight for the client. We use our expertise in statistics to develop insights and understanding.
A project may be a multi-phased engagement. For example, an initial phase may involve the prototyping of a predictive model and data analytics pipeline, and a subsequent phase may conduct repeated analyses using new data or novel features.
Typically, we work with clients on a per project basis (using a fixed price agreed to in advance). In some instances, we work on a "time basis" using pre-paid Time Blocks (e.g., 10h or 20h per week).