Software engineer at IBM building developer tools, data APIs, and agent infrastructure.
PhD in Statistics & Computer Science. Former forensic statistics researcher. R package author and maintainer.
I build tools that help people see their data, and agents that help them act on it.
Right now I work on IBM's Developer and Operator Experience team, mostly around Langflow - one of the most-starred open-source AI projects - along with the Astra DB Data API, document ingestion, and the practical edges of agent tooling: SDKs, MCP integration, model-provider abstractions, tool-use policy, and production ergonomics.
Before that I was at DataStax, Daisi Technology, and Omni Analytics Group. Before that, I did a PhD at Iowa State on statistical methods for matching fired bullets to the guns that fired them, using 3D surface scans, reproducible analysis pipelines, and statistical visualization.
- Langflow core - SDK and Flow DevOps API work, MCP auth resilience, provider cleanup, and the small reliability fixes that make open-source tools feel solid.
- Agent safety and tool policy - contributing to ToolGuard and bringing policy-aware tool use into Langflow.
- Document parsing and ingestion - integrating IBM's Docling into Langflow and Astra DB workflows so PDFs, Office docs, scans, and slides become useful context without bespoke glue code.
- Forensic statistics projects - occasional work on bullet-matching tools,
.x3ptopography, and browser-native visualization like Bullet Lens.
| Project | What it is |
|---|---|
| Langflow | Open-source visual framework for building agentic AI applications. I am a core maintainer. |
| Astra DB Data API | HTTP/JSON API surface, Python client work, ingestion services, and vector/hybrid-search workflows. |
| Docling integration | Bringing IBM's document parser into Langflow and the Astra DB ingestion stack. |
| Aeroza | Weather-intelligence platform - real-time radar, NWS alerts, geospatial queries, nowcasting, and forecast verification. |
| bulletr / bulletxtrctr | R tooling for bullet-striation extraction, matching, and forensic statistics research. |
| discreteRV | R package for teaching and manipulating discrete random variables, published in the R Journal. |
| OmniacsDAO | Open-source data-science tooling for Numerai, DeFi analytics, and experiment workflows. |
Languages
Web & API
Data & infra
AI & agents
- Automatic Matching of Bullet Land Impressions - Annals of Applied Statistics (2017). doi:10.1214/17-AOAS1080
- Designing Modular Software: A Case Study in Introductory Statistics - Journal of Computational and Graphical Statistics (2017). doi:10.1080/10618600.2017.1386568
- Manipulation of Discrete Random Variables with discreteRV - The R Journal (2015). RJ-2015-005
- Biomathematical Description of Synthetic Peptide Libraries - PLOS ONE (2015). doi:10.1371/journal.pone.0129200
- Putting Down Roots: A Graphical Exploration of Community Attachment - Computational Statistics (2018). doi:10.1007/s00180-017-0735-1
Full list: erichare.me/publications
- PhD, Statistics & Computer Science - Iowa State University (2017) · dissertation: Statistical Methods for Bullet Matching
- MS, Statistics - Iowa State University (2014)
- BS, Statistics & Computer Engineering (dual) - University of Washington (2012)
- AI agent tooling, MCP, and tool-use policy
- Open-source developer experience
- Document parsing, ingestion, and retrieval workflows
- Statistical graphics and reproducible research
- R and Python packaging
- Forensic statistics and teaching through software
- Website: erichare.me
- Projects: erichare.me/projects
- Publications: erichare.me/publications
- CV: erichare.me/cv
- Now: erichare.me/now
I am based in Lake Stevens, Washington. Best way to reach me is ericrhare@gmail.com.