MANIL MEHTA

FULL-STACK DEV • AI ENGINEER/RESEARCHER • LLMS

TRAVEL • MUSIC • VOLLEYBALL

data science student at uc davis open to summer 2026 internships/research

ABOUT ME

Who I Am

I'm a Data Science student (currently a freshman at UC Davis) who likes building things that actually have real-world applications and work outside of demos. I've been coding since 2020, mostly because I enjoy understanding how systems behave under real constraints (eg. latency, bandwidth, scale, etc.).

Beyond Code

Love volleyball (on my college's Division 1 team), play the piano, and spend my free time building things I think are cool.

Skills

PythonJavaJavaScriptSQLHTML/CSSAWSGCPDockerKubernetesGitMongoDBNode.jsExpressReactFlaskTailwind CSSPandasScikit-LearnKerasPyTorch

EXPERIENCE

Undergraduate Researcher

Data Driven Change @ UC Davis

Jan 2026 – Present • UC Davis

  • • Building and evaluating classification and clustering models to segment donor populations and predict volunteer activation probability
  • • Performing feature engineering on high-dimensional behavioral and demographic datasets to improve model performance and interpretability
  • • Implementing model validation workflows using cross-validation and performance metrics (AUC, precision-recall, F1) to ensure reliability in real-world deployment

Machine Learning Developer

AI Student Collective

Jan 2026 – Present • UC Davis

  • • Built an AI-powered chat interface that enables students to query professor performance data (e.g., course difficulty, grading trends, exam formats) using natural language
  • • Productized the platform for UC Davis students, driving real-user adoption and validating demand for AI-assisted academic planning tools

Digital Systems Manager

Men's Club Volleyball @ UC Davis

Sep 2025 – Present • UC Davis

  • • Developed and maintained a production web platform for UC Davis Men’s Volleyball, managing site updates, performance optimization, and responsive UI components
  • • Integrated e-commerce functionality to support merchandise sales, payments, and inventory management
  • • Designed and deployed digital content pipelines across social platforms to increase audience reach and fan engagement

Software Engineering Fellowship

Palantir Technologies

Nov 2025 – Present • Remote

  • • Built UAV batch/streaming ML pipelines in Palantir Foundry with telemetry normalization and lineage-tracked inference outputs
  • • Deployed YOLOv8/ResNet-50 inference and FAISS-backed RAG pipelines on Palantir AIP for maintenance triage
  • • Instrumented ML observability pipelines, tracking drift, data integrity, and inference reliability with Evidently inside Foundry

Software Developer

Project RISHI @ UC Davis

Sep 2025 – Present • Davis, CA

  • • Built a React Native mobile application for medication adherence with scheduled notification workflows and persistent user state
  • • Implemented notification scheduling, cancellation, and update logic to handle user changes, missed reminders, and lifecycle events
  • • Designed privacy-aware reminder timing algorithms to reduce stigma and improve medication adherence in healthcare settings
  • • Collaborated with healthcare stakeholders to convert clinical requirements into production-ready mobile features

Data Science Extern

Amazon

June 2025 – August 2025 • Remote

  • • Built an onboarding review triage system using MiniLM embeddings and LightGBM to rank early employee attrition risk
  • • Preprocessed 5K+ employee reviews with pandas and scikit-learn; improved precision@10 by 18% over baseline
  • • Applied SHAP to explain LightGBM rankings, highlighting influential tokens in the top 10% of high-risk flagged reviews

Co-Founder — Program Director

STEMplicity

Jun 2023 – May 2025 • San Francisco, CA

  • • Architected and scaled program delivery systems supporting 1,000+ users, optimizing scheduling and operational workflows
  • • Designed CS curricula and assessment frameworks focused on algorithmic thinking and computational problem-solving
  • • Built and ran CS Bee, a large-scale technical evaluation program assessing CS fundamentals across cohorts
  • • Led and managed cross-functional teams, owning execution and end-to-end delivery

PROJECTS

WIP

ATLAS

FastAPI, Python, PostgreSQL, Docker, Whisper

Built a multimodal reasoning pipeline for event reconstruction from text, image, video, and audio inputs. Developed an agentic memory pipeline that tracks user context and adapts next-day lessons based on learning history.

BE-FAIR-Proposal

Python, Matplotlib, Latex

Extension of Gupta et al.'s BE-FAIR framework: rapid bias screening methodology using relative error metrics to identify healthcare AI disparities across race/ethnicity and social vulnerability.

SpeechMe

Python, Next.js, Hugging Face STT, LLMs, MongoDB

Developed a multilingual emergency response agent converting unstructured speech into structured MCP dispatch actions. Integrated DeepL translation, Perplexity classification, and Apify retrieval for location-aware autonomous routing.

WIP

MedWise

Next.js, Flask, Neo4j, Pinecone, Hugging Face

Developed a graph-based RAG system linking 1,000+ drugs with embeddings in Pinecone for generics and interaction queries. Integrated Neo4j graph traversal with Hugging Face rerankers to improve retrieval precision and clinical interpretability.

WIP

Wizard

JavaScript, Node.js, TypeScript, Docker, Postgres, Shell scripting

Built Wizard, a CLI debugging tool with sandboxed error reproduction, syscall tracing, and structured root-cause analysis. Added automated fix workflows using semantic error classification, environment diffing, and safe command execution.

View Full Notes →Download Resume →

GET IN TOUCH

Looking for internship opportunities for Summer 2026

mmmehta@ucdavis.edu