Currently PM Fellow at Product Space

Rehan
Hashim.

Data science at Illinois. Lately I’ve been working on knowledge graphs for LLM agents, RAG pipelines, AI product strategy, ROI models for AI tools

See my work LinkedIn or press to explore

Strategy✦Requirements✦Prioritization✦ROI modeling✦Competitive analysis✦User research✦Python✦SQL✦Java✦R✦Neo4j✦RAG✦LLM agents✦Data pipelines✦Git✦Jira✦Azure DevOps✦Cursor✦Claude Code✦Google Cloud✦Strategy✦Requirements✦Prioritization✦ROI modeling✦Competitive analysis✦User research✦Python✦SQL✦Java✦R✦Neo4j✦RAG✦LLM agents✦Data pipelines✦Git✦Jira✦Azure DevOps✦Cursor✦Claude Code✦Google Cloud✦

01 / Experience

Where I’ve been working

May – Aug 2026 · a Deloitte company · Chicago

Software Engineering Intern @ Wavicle Data Solutions

Worked on Nova, the first product in Wavicle's Meridian suite: an AI-native data intelligence platform that structures enterprise data in a Neo4j knowledge graph (Atlas) for retrieval by LangGraph and Claude agents (Compass).

Built the Redshift metadata ingestion script in Python, with a simulation mode so the team could test the pipeline end to end without live client data.
Built the Atlas loader that writes metadata into Neo4j as typed nodes and relationships using MERGE, keeping repeated loads free of duplicates.
Designed a dual-agent retriever/reasoner pattern for Compass and recommended domain partitioning to improve agent retrieval accuracy.
Identified semantic drift detection as a defensible differentiator, and rebuilt the platform's testing documentation ahead of the testing phase.

Python
Neo4j
SQL
LLM Agents
LangGraph

02 / Projects

Things I’ve built

012026
Winner · Caterpillar Track
CAT Time Machine
A time machine for your jobsite.
UIUC Product Hackathon · Hack to the Future · Team of 4
Replay any recorded day on a mining site in a 3D spatial simulation, ask “why was production down?” in plain English, and pull up the full context of the incident. Built for Caterpillar's challenge to give machines and jobsites a persistent memory.
- Scoped the product around protest and strike roadblocks that halt mine transport routes, a scenario that has cut major mine output to 30% of capacity.
- Built a 3D replay of past site operations paired with an AI assistant that retrieves the affected trucks, weather and production data behind a disruption.
- Exposed incident memory to external AI tools through an MCP connector so the next shift can reuse lessons from past disruptions.
- AI Agents
- MCP
- 3D Simulation
- Data Viz
Devpost Demo video GitHub

03 / About

Beyond the resume

Rehan Hashim standing on a sand dune — Rehan Hashim · Illinois ’28

Hi, I’m Rehan

I'm studying Information Sciences and Data Science at the University of Illinois Urbana-Champaign. I like the space between building the system and deciding what the system should be.

This summer I built knowledge-graph tooling for LLM agents at Wavicle Data Solutions. Now I'm a PM fellow at Product Space, consulting on AI commercialization with EntreCORPS, and helping student founders through Founders Illinois.

Education · 2025 – 2028

University of Illinois Urbana-Champaign

B.S. Information Sciences & Data Science

Database Systems, Statistical Analysis, Linear Algebra, Data Structures & Algorithms

Psi Eta Mu Professional Data Science Fraternity

$600K

projected savings in the ROI model I built for Ask jAIne

50+

founders in my Founders Illinois speaker pipeline

universities in the field trial I'm designing

acceptance rate into Product Space

Toolkit

Product: Strategy
Requirements
Prioritization
ROI modeling
Competitive analysis
User research
Technical: Python
SQL
Java
R
Neo4j
RAG
LLM agents
Data pipelines
Tools: Git
Jira
Azure DevOps
Cursor
Claude Code
Google Cloud

Off the clock

Man United
Poker
Traveling
Metalcore
80s music
Berlin Global Leadership Summit delegate

Where I’ve been working

Software Engineering Intern @ Wavicle Data Solutions

Things I’ve built

CAT Time Machine

Beyond the resume

University of Illinois Urbana-Champaign