Work

[SCREENSHOT — Discover / Dashboards]16:10 crop
elastic · director of engineering · 2019 — presentactive

Kibana Application Experience

PROBLEM
The analytics surfaces most Elastic customers open first were spread across separate teams with separate ideas about what good looked like.
APPROACH
Built one group of four teams under three managers — Discover, Dashboards, the Elastic design system, and the ML, vector search and agentic interfaces — with a shared bar for craft and a design system everyone actually uses.
OUTCOME
45 engineers owning roughly 70% of Kibana customer usage. [ADD ONE HARD RESULT — adoption, latency, satisfaction, whatever you can point at.]
ReactTypeScriptDesign systemsOrg design
[DIAGRAM — practice rollout]or a photo from a training session
elastic · founder & leadactive

Kibana Applications AI Working Group

PROBLEM
Agentic development was happening in pockets — a few people getting real leverage, everyone else guessing.
APPROACH
Founded a working group to turn those pockets into a practice: shared patterns, a point of view on where judgment still belongs to humans, and training built for engineers who'd rather see it than read about it.
OUTCOME
150+ engineers trained and running the practice. [ADD A SECOND RESULT — cycle time, adoption rate, or what teams now do differently.]
Agentic workflowsEnablementCode review
[PRODUCT SHOT — booking flow]swap for a live demo embed later
independent · solo founderin progress

Workforce travel platform

PROBLEM
Crews deployed to oil fields and datacenter builds get booked through spreadsheets, group texts and whoever happens to answer the phone.
APPROACH
A B2B product for booking hotels and coordinating travel logistics for industrial workforces at remote job sites — designed, built and shipped solo.
STATUS
[CURRENT MILESTONE — e.g. private beta with N operators, or first booking processed.]
PythonReact[DB][HOST]
[GAMEPLAY CAPTURE]animated gif works here
side practice · developershipped

Live multiplayer experiences on Roblox

PROBLEM
Staying hands-on with real-time systems — state you don't control, players who find every edge, and no room for a slow tick.
APPROACH
Server and client scripting, round flow and team assignment, collision handling, RemoteEvents and DataStore persistence — built and debugged end to end.
OUTCOME
[PLAYS / SESSIONS / WHAT YOU LEARNED — plus a link to the experience.]
LuauRemoteEventsDataStore
$ gh repo list --limit 3 all repos →