Most healthcare data leaders aren’t struggling with analytics.They’re struggling with inherited chaos — 10+ years of conflicting definitions, EMR exports nobody trusts, and leadership demanding answers from numbers that don’t align.

I partner with the people living that world every day: VPs of Analytics, Directors of IT
and cross-functional data leaders who are drowning in Excel reconciliation, political misalignment, and dashboards that spark more debate than decisions.

If your finance, operations, and EMR teams all have different definitions of “Net Revenue,” I fix that.

I build the data foundations that make AI possible — and trustworthy.

DM me "AI in Healthcare" if you're ready to get your data foundation in order.

Christian Steinert

Finch AI

- Collaborated with stakeholders, project managers, and other data scientists, to design data collection and LLM development efforts to improve NLP models in the backend of publicly facing software solutions.

- Led AI teams of data scientists end-to-end from data curation to model deployment to update to our model offerings, accomplishing a 25-point increase in model performance and heightened customer satisfaction.

- Incorporated LLMs and generative AI models, such as GPT-3.5, GPT4, Mistral, LLAMA, etc., into product offerings utilizing prompt engineering, RAG, and vector stores like Chroma DB, leading to increased work throughput.

- Utilized AWS EC2 instances and Linux to automate training and evaluating pipelines for NLP and NER models.

- Sourced, validated, and managed high-quality data for model training and development.

- Mentored junior data scientists providing them technical guidance resulting in improved cross-team learning and collaboration.

Data Scientist

University of Colorado Boulder

- Co-created an error analysis Python library to aid researchers’ hate speech detection natural language processing (NLP) models so that more ethical AI systems can be built.

- Researched the most current machine learning, neural network, and deep learning models to understand where limitations are in current research and design the library to facilitate in understanding where such models fail linguistically.

- Documented, co-wrote, and submitted an accepted paper to EACL 2023 about the Python library and appropriate use cases.  The library can help models improve by at least 5 percent.

NLP Researcher

Brandeis University

- Built, trained, and fine-tuned machine learning models in R and Python via traditional statistics-based data science techniques, Bayesian logistic regression, feature engineering, and unsupervised clustering algorithms, saving $50,000 in project costs and shaving 11 months off research and development time.

- Defined project goals to stay in close alignment with business needs and goals to timely deliver models, insights, and data visualizations to customers. 

- Automated repetitive workflows and processes to swiftly iterate models and their reliability, as well as simplifying and streamlining the supporting code bases.

- Clearly communicated technical concepts to key business stakeholders with data-driven insights and A/B testing to influence the direction of model development.

- Ensured data governance by managing the Linux machines which contained model data and updating documentation.

Luminoso Technologies, Inc.

- Leveraged Japanese linguistic knowledge, drastically improving NLP sentiment analysis by over 50% for Japanese text data, leading to the product selling to a Japanese audience. 

- Articulated and arranged relevant Japanese to English translations, thereby refining and analytic processes by 20%.

NLP Intern

My name is Jay Seabrum and I am an experienced and passionate data scientist driving innovation in the field of machine learning and AI with 4 years of working in small teams, solving complex and real-world end-to-end problems.  Prior individual contributions include: over $50,000 saved in costs via innovative and efficient processes, staying up to date with advancements in AI and ML, leading teams in developing machine learning models, and a decrease in model iteration and experimentation time. Aiming to further apply expertise in a senior or mid-level data scientist role, or more leadership oriented roles.  Open to relocation.

You can contact me in one of three ways:  (1) schedule a meeting on calendly.com/seabrum , (2) send a message with your connect request, or (3) send an email to seabrum.x@gmail.com and I'll be sure to get back with you.  Please note that (1) I do not accept connect requests without a message attached to them and (2) I am not interested in pure software engineering roles. Thanks!

#datascience #datascientist #nlp #machinelearning #natural #language #processing #ai

taplio

The Battle

Christian Steinert

Xajavion Seabrum

Followers

Posts

Engagements

Likes

Ready to grow your LinkedIn presence?