Home Current Projects Personal Blog
Headshot of Mohan Kolla

Mohan Kolla

Data Scientist & Software Engineer

mokolla21@gmail.com

LinkedIn

github.com/mohan-kolla


Skills

Programming Languages
PythonJavaScriptSQLJavaC#RMATLABHTML/CSS
Frameworks & Libraries
ReactNode.jsExpress.jsBootstrapPyTorchScikit-learnPandasNumPyStatsmodelsMatplotlibSeabornGradio
AI & Machine Learning
Retrieval-Augmented Generation (RAG)Multimodal RetrievalLLM Application DevelopmentVision-Language ModelsPrompt EngineeringNeural NetworksMachine Learning AlgorithmsModel Evaluation & BenchmarkingRegression & Statistical Modeling
Data Engineering & Analytics
ETL/ELT PipelinesData IngestionData WarehousingRelational Database DesignGeospatial SQLArcGIS (arcpy)Power BITableauData Visualization
Cloud, Tools & Platforms
AWSDockerGit/GitHubLinuxREST APIsOpenAI APIHugging FaceGPU / HPC ComputingcondaWeights & Biases
Practices
AgileScrumSystem DesignPipeline AutomationTesting & ValidationTechnical DocumentationTechnical Presentations

Languages

English

Telugu

Spanish



Engineering & Research Experience

Data Systems Developer (Part-Time) / Bawarchi Biryanis
2024 - Current

Family-owned restaurant, Miami, FL

  • Designed, built, and maintain the restaurant's inventory and sales data system, in active use by the business since 2024.
  • Built Python ETL pipelines that load spreadsheet-based inventory and sales records into a SQL data warehouse.
  • Built a Power BI reporting dashboard the business uses to model food costs, plan catering orders, and reduce food waste.
  • Support seasonal demand swings, with the heaviest use during holiday catering peaks.

Graduate Research Assistant / Florida International University, Robotics & Spatial AI Lab
August 2024 - September 2026
  • Built the data ingestion and retrieval backend for MHyCube-RAG, a multimodal retrieval-augmented generation system for compound flood reasoning.
  • Engineered pipelines integrating NASA IMERG satellite precipitation, NOAA MRMS radar rainfall, and NOAA CO-OPS tide gauge data into a shared spatiotemporal data store, processing 50GB+ of structured and unstructured data.
  • Built an evaluation benchmark and ran a six-way retrieval comparison (hypercube, dense, BM25, FLARE, GraphRAG, and fusion multimodal), validating 96% correctness and 94.7% completeness against ground truth.
  • Deployed the MarineGPT vision-language model on shared A100 GPU servers and built pipelines that capture snapshots from 10 ocean livestreams every 5 minutes and caption them automatically, running unattended.
  • Reused the system's architecture for a second application in fisheries (FishRAG). Code: FIU-MoRA-Lab/Hypercube-RAG-FIU.

Research & Presentations

MHyCube-RAG: An Integrated Multimodal RAG Architecture for Compound Flood Reasoning
IEEE ICTAI 2026 (conference paper submission)

Shows that sensor-grounded multimodal retrieval answers predictive spatiotemporal flood questions that text retrieval alone cannot.


Poster Presentation, i-GUIDE Conference
August 2026, University of Illinois Chicago

Presented the MHyCube-RAG framework at the NSF-supported Institute for Geospatial Understanding through an Integrative Discovery Environment conference.

Education

Master of Science in Data Science and Artificial Intelligence
2023 - 2026

Florida International University, Miami, FL


Bachelor of Science in Biochemistry
2017 - 2021

University of Florida, Gainesville, FL

Certifications

  • AWS Certified Cloud Practitioner
  • Databricks: AI in the Flow of Data Engineering
  • PCEP: Certified Entry-Level Python Programmer
  • Microsoft Azure Data Engineer Associate (in progress)

Other Experience

  • Academic Tutor, On Point Tutoring (November 2023 - Current): one-on-one tutoring in calculus, regression modeling, machine learning, and programming.
  • Medical Transcriptionist, Quest Diagnostics (October 2022 - May 2023)
  • Medical Assistant, Mendez Medical Center (August 2021 - March 2022)