adam@hub:~$ ./init --scene rooftop
[ ok ] tokyo-night palette loaded
>

adam_rodriguez

SENIOR SOFTWARE ENGINEER · AI SYSTEMS
builds and runs the ai systems a business depends on daily

i build systems that keep running after i walk away. that is the whole thing, in both halves of my work.

the day job is enterprise health-insurance data at eight million files a day — kafka-driven pipelines with real-time observability, data-cleanliness and value-validation passes that stop bad values before anything downstream sees them, a react/typescript document viewer handling pdf, tiff and text behind one search surface, and three applications migrated off websphere onto spring in three months. large systems, real constraints, no room to hand-wave.

the other half is smaller and moves faster. a content engine that has run daily in production for five months — a brief goes in, finished on-brand content comes out. a lead machine that sourced 11,795 prospects and separated the warm list from the cold one. same instinct at a different scale: find the process someone repeats by hand, and build the thing that does it every day without being asked.

what i am looking for now is the version of this with more surface area — owning the ai systems a company depends on, not just the service that sits underneath them.

registersenterprise · small business
in the hub4 systems you can open
works inkafka · java · distributed systems · observability · react · typescript
basedspace coast, florida
8M
files a day through production pipelines
3 apps
migrated off websphere onto spring · in 3 months
11,795
prospects sourced · deduped and segmented
5 mo
content engine · running daily in production
featured6
the archive1
florida blueenterprise
  • data_at_scalehealth-insurance data pipelines — 8M files/day, kafka-driven stages, real-time message-flow observability · MAS
  • document_frontendmodular react/typescript viewer — pdf · tiff · text, citation highlighting, lexical + semantic search · ECMS
  • ai_integrationmulti-agent pipelines reconciling metadata against document content, at enterprise scale · MAS
  • java_migrationwebsphere → spring migration — three applications, one frontend and two backends · PAISdelivered in 3 months against a 1–2 year estimate~$1m in avoided hiring cost (est.) — built in-house instead of bringing on a java contractor
8 million files a day · in production
jodi jones studiosmall business
  • content_enginea brief goes in · finished on-brand content comes out · every day
5 months · in production · running daily
manny rodriguez photographysmall business
  • lead_machineicp → sourced → enriched → drafted · repeatably
11,795 prospects sourced · warm list separated from cold
tell me what your team is trying to build · i’ll tell you where i’d fit
systems for businesswhat i build · and how we'd start