Software engineer and statistician.
Core Langflow maintainer at IBM; creator of Verity and Aeroza.
Work · Research · Writing · About · CV · Now
I'm a software engineer at IBM on the Developer and Operator Experience team, where I'm a core maintainer of Langflow. My work sits at the boundary between the clean API a developer wants and the messy production system that has to keep working: agent tools, MCP, authentication, model providers, retrieval, packaging, SDKs, and release pipelines.
My earlier work focused on statistical evidence. I earned dual undergraduate degrees in Statistics and Computer Engineering from the University of Washington, then a PhD in Statistics and Computer Science at Iowa State. My dissertation developed reproducible methods for comparing 3-D scans of spent bullets; that research now has a second life in Verity. I came to IBM through its acquisition of DataStax, after earlier roles at Daisi and Omni Analytics.
The common thread: build systems that show their reasoning, state, uncertainty, and failure modes—and ship the implementation so other people can inspect it, run it, and challenge it.
Systems built to hold up under scrutiny.
CORE MAINTAINER · 2024 — PRESENT
The open-source visual framework for building AI agents. I help shape Langflow at IBM, with a focus on agent tooling, MCP, the SDK, and the production details that make open-source software dependable.
Explore the work · Repository · Merged PRs
CREATOR · 2026 — PRESENT
I'm building Verity because forensic conclusions should expose uncertainty, not hide it behind a binary match. It reports calibrated evidentiary strength from 3-D scans of bullets, cartridge cases, and toolmarks.
Explore the work · Live platform · Repository
CREATOR · 2026 — PRESENT
I'm building Aeroza to make weather data queryable—and forecasts accountable to what actually happened. It brings live radar, geospatial queries, streaming APIs, signed webhooks, and probabilistic nowcasting into one open platform.
Explore the work · Live platform · Repository
| Project | What it is |
|---|---|
| AI Workbench | A self-hosted workspace for building, inspecting, and operating retrieval-backed AI applications on Astra. Case study → |
| Astra DB Data API | The HTTP/JSON data plane behind Langflow, ingestion, and the astrapy Python client. |
Notes on evidence, agents, and uncertainty.
| Date | Writing |
|---|---|
| Jun 2026 | Verity: weighing the marks, not calling the match — calibrated likelihood ratios for forensic surface comparison. |
| May 2026 | Aeroza: weather, but queryable — weather infrastructure with public forecast verification. |
| Apr 2026 | Bullet Lens: forensic topography in the browser — interactive 3-D inspection of .x3p bullet scans. |
The work behind Verity began with my dissertation and Automatic Matching of Bullet Land Impressions (Annals of Applied Statistics, 2017), plus open R packages for 3-D surface analysis and statistical comparison.
Read the research story · Publications · ORCID
Let's build things that stand up to scrutiny.
I'm always up for a conversation about AI agents, developer platforms, statistical graphics,
or forensic statistics.
Email ·
LinkedIn ·
erichare.me