1st
Best Overall — HBA×CSBA Hack Day
Keystone: a LangGraph/Gemini multi-agent real estate platform, built in 5 hours.
01 — Intro
I'm Rayhan — B.S. Computer Science and Statistics & Data Science at UT Austin. I ship production AI: Vertex multi-agent platforms at Humana, a $300K-backed agentic marketplace at Krowe, RAG systems at PGA, and genomic ML on TACC HPC — always with CI/CD, evals, and measurable impact.
4.0
GPA
2
B.S. degrees
8
Roles
2×
Hackathon wins

02 — Approach
End-to-end ownership, from first commit to production — founding engineer at Krowe, and enterprise AI deployed to every Humana L2 team.
Evals, telemetry, and statistical validation so features stay reliable after launch — the habit behind 40% fewer data errors in genomics research.
Clear APIs and reproducible workflows make the next version cheap. Latency and correctness are hard constraints.
Current focus
03 — Experience
Summer 2026 → Present
Software Engineer Intern
Humana · Louisville, KY
72%
Faster certification cycles
18
Python automation scripts
Deployed AI products to every L2 team through a Vertex AI multi-agent platform with Azure DevOps CI/CD — and cleared a 90-day backlog for 13 engineers.
May → Aug 2025
Machine Learning Engineer Intern
The PGA of America · Dallas, TX
12
ML systems shipped
500+
Engineers adopted the tooling
RAG pipelines in Python/FastAPI/Docker with CI/CD alongside 14 ML engineers. Standardized Docker tooling and cleared integration blockers 3 weeks early.
Aug 2023 → Present
Founding Software Engineer
Krowe Technologies · Dallas, TX
$300K
Raised
23
Startups
Modular agentic AI platform and developer marketplace on React/Next.js and FastAPI, letting AI developers sell custom agentic tools to SMB clients.
Nov 2025 → Present
Undergraduate Researcher
UT Center for Computational Biology & Bioinformatics · Austin, TX
40%
Fewer data errors
HPC
Pipelines on TACC clusters
Python ML pipeline identifying genomic mutation sites in time-series data, with automated statistical validation and reproducible Git workflows on Linux.
May → Aug 2024
Software Engineer Intern
Nixar Solutions · Allen, TX
18%
Conversion lift
533%
Interaction growth
Five-figure revenue shipping React/Next.js features with multi-agent AI pipelines. A/B tests and KPI dashboards that boosted reach 582%.
Aug 2025 → Present
Technology Consultant
Technology Consulting Group · Austin, TX
$120K
Raise enabled for XFund
27
States mapped
Agentic investment-diligence system on the Claude API for XFund ($190M fund), plus a 27-state geospatial pipeline and dashboard for Well Water Finders.
Jul 2025 → Present
Software Engineer
Texas Convergent · Austin, TX
35%
Throughput gain, AMOS Labs
Live
HealthKit athlete dashboards
TypeScript recursive AI agent for AMOS Labs (Capital Factory), and a React Native/Node.js app streaming Apple HealthKit biometrics to live athlete dashboards.
Jul 2025 → Present
ML & Software Developer
Business Analytics Association · Austin, TX
1st
Of 150 teams, T-Mobile Hackathon
140
Members served
NumPy/Keras ML pipeline for a 140-member analytics club, plus a React/Supabase member portal tracking engagement across 8 monthly events.
04 — Wins
1st
Keystone: a LangGraph/Gemini multi-agent real estate platform, built in 5 hours.
1st/ 150
NumPy/Keras ML pipeline with the Business Analytics Association.
$300K
Raised
Founding engineer, Krowe
$120K
Raise enabled
XFund diligence, TCG
1,000+
Students served
Campus Connect AI
72%
Faster cert cycles
Humana
05 — Education & research
The University of Texas at Austin
B.S. Computer Science
+ B.S. Statistics & Data Science
Class of May 2028 · Dual degree
4.0
Cumulative GPA
Research
Nov 2025 → Present
UT Center for Computational Biology & Bioinformatics
Identifying genomic mutation sites in time-series data with a Python ML pipeline on TACC HPC clusters — automated statistical validation cut data errors 40%.
06 — Toolkit
Shipped with
Python
9 places · hover or tap a tool
End of file — Next: contact
Open to software and ML engineering roles — especially agentic AI platforms, RAG infrastructure, full-stack product, and data-intensive systems.