86°35′W · 34°43′N · HUNTSVILLE
Research Engineer (ML & Backend) — NASA IMPACT
Research Engineer (ML & Backend) at NASA IMPACT. Production-grade AI and cloud-native applications in Python, AWS and React: RAG systems, data pipelines and automation platforms.
01 — Selected work
Three personal prototypes from 2026. Each case study shows the problem, the decisions, what was measured and what still fails.
All projectsA personal prototype that builds Grafana dashboards from a sentence, inside Grafana: an app plugin in TypeScript and Go, and a Python agent that plans the panels, builds them with typed builders and checks that every saved panel returns data.
A personal prototype of a chat agent that turns plain-English requests into searches of NASA’s commercial smallsat catalogue, resolving vague wording through a catalogue graph and keeping a map panel in sync with every tool call.
A solo build for the IBM BoB Hackathon that turns one sentence into a Sentinel-2 image chip, scoring cloud over the area actually asked about and applying the reflectance offset most scripts miss. Two scenes show why.
02 — The Lab
Ask for Earth-observation data in plain words, a place and a time. A small model plans the search with four read-only tools; ordinary code checks every argument and builds every request.
Find high-resolution radar (SAR) scenes over the San Francisco Bay Area from the second half of 2025. I need imagery that sees through fog.
search_catalogue
5 candidate collections: capellaspace, iceye, mda, umbra, airbus
resolve_place
→ bbox [-122.7, 37.2, -121.7, 38.1]
search_items
0 scenes found
search_items
10 scenes found
Umbra returned scenes: UMBRA-07 on 2025-12-23 and 2025-11-27, UMBRA-08 on 2025-08-09. SAR images through fog and cloud, so no cloud-cover value applies. Iceye, MDA and Airbus were not searched.
03 — Stack, in cross-section
Read it like a borehole: the bedrock I’ve worked in longest sits at the bottom, the newest layer at the surface. Every layer is something I’ve shipped with.
Layer thickness follows years of use, not skill level.
LangGraph · LangChain · Pydantic AI · MCP
Since 2024 →
RAG · Weaviate · ChromaDB · pgvector · knowledge graphs
Since 2024 →
STAC catalogs · Airflow · rasterio · Sentinel-2 · PostgreSQL
Since 2024 →
S3 · EC2 · RDS · Lambda · Step Functions · ECS · SageMaker
Since 2019 →
Django · FastAPI · PyTorch · scikit-learn · pandas
Since 2019 →
04 — 03 — Experience
042024 — Now
Research Engineer (AI & Backend)
Huntsville, AL · part-time
Data Acquisition Request System
032023 — 2024
Software Engineer (ML & Backend)
Palo Alto, CA · remote
Houzz Moodboard
022021 — 2023
Software Engineer (ML & Backend)
Seattle, WA · remote
Trading Platform
012019 — 2021
Software Engineer (Full-Stack Development)
Hessen, Germany · remote
ListInfo
05 — Writing
Moon, Mars and Earth data in one traceable model: what I built, the problems I hit, and how I ran several coding agents without losing track of the work.
