// booting portfolio
Manav
Prakash
00%

Computer science undergraduate at VIT Vellore, working across AI automation, software engineering, and data analysis. Happiest on a team, on a problem nobody has solved yet.

scroll
drag them around
01

About

Vellore, Tamil Nadu

I’m a Computer Science and Engineering undergraduate at VIT Vellore who enjoys turning ideas into practical, impactful technology. From building software and exploring cloud-based systems to experimenting with AI, cybersecurity, and automation, I’m driven by the challenge of solving real-world problems with clean, creative solutions.

I believe the best technology is built through curiosity, collaboration, and continuous learning. Whether I’m debugging a project, designing an intelligent workflow, or learning a new tool, I enjoy working with people who challenge assumptions and turn ambitious ideas into reality.

Beyond code, I’m passionate about understanding how technology can make systems smarter, more secure, and easier to use. I’m constantly expanding my skills through hands-on projects, problem-solving, and exploration of emerging technologies—aiming to grow into an engineer who builds solutions that are not only technically strong, but genuinely useful.

0.00

Cumulative GPA across six semesters

0

Engineering internships shipped to stakeholders

0

Products and research projects built end to end

0+

Certifications in cloud and machine learning

02

Selected work

7 projects, open any row
03

Experience

2025 — 2026
Jun — Aug 2026Bengaluru, Karnataka

Technology Development Program Intern

Optum Global Solutions

  • Partnered with healthcare business stakeholders to find automation opportunities, scope proofs of concept, and translate business needs into technical requirements for medical document workflows.
  • Built Azure and Python extraction pipelines using agentic LLM frameworks, converting unstructured medical documents into structured records with higher accuracy and less manual review.
  • Designed end-to-end ingestion, validation, and routing so claims and admin decisions could be made faster.
  • Wrote Playwright end-to-end tests against the workflow PoCs, catching regressions before stakeholder demos.
  • Worked across time zones with US stakeholders, and measured pilot impact — manual effort saved, turnaround time — to guide rollout decisions.
PythonAzureAgentic LLMPlaywrightDocument AI
Aug — Sep 2025Vellore, Tamil Nadu

AI / ML Intern

CiviLytix

  • Built lightweight vision models for in-car detection of road quality issues.
  • Optimised those models for real-time inference on low-spec automotive hardware.
  • Contributed to data pipelines that push road quality metrics to central systems for mobility analytics.
PyTorchTensorFlowYOLOONNXOpenCVEdge inference
04

Toolkit & credentials

Updated 2026

Languages

  • Python
  • Java
  • C / C++
  • JavaScript
  • TypeScript
  • SQL
  • R
  • Arduino C++

AI & machine learning

  • Agentic LLM frameworks
  • RAG
  • Vector databases
  • PyTorch
  • TensorFlow
  • scikit-learn
  • Hugging Face
  • spaCy

Vision & imaging

  • OpenCV
  • CLAHE
  • Dark Channel Prior
  • Multi-scale fusion
  • YOLO
  • ONNX
  • Edge inference

Security & cryptography

  • ML-KEM-768
  • ML-DSA-65
  • Scapy
  • mitmproxy
  • CVE analysis
  • Penetration testing
  • Classical cryptography

Data & analysis

  • pandas
  • NumPy
  • Matplotlib
  • Feature engineering
  • Data cleaning
  • SQL
  • Pivot tables

Cloud & backend

  • AWS Lambda
  • Step Functions
  • Textract
  • DynamoDB
  • Azure
  • GCP
  • Firebase
  • Flask
  • Node.js

Web & interfaces

  • React
  • Express
  • Tailwind
  • Bootstrap
  • Streamlit
  • REST APIs

Testing & tooling

  • Playwright
  • Git
  • VS Code
  • PyCharm
  • Visual Studio
  • Arduino IDE

Foundations

  • Data structures
  • Algorithms
  • OOP
  • Networking
  • Embedded systems
  • IoT protocols
  • Technical writing

in use on a project I’m building right now

Jul 2025Oracle Cloud Infrastructure Generative AI ProfessionalOracle · 1Z0-1127-25
Jun 2025Oracle Cloud Infrastructure AI Foundations AssociateOracle · 1Z0-1122-25
May 2025Machine Learning SpecializationDeepLearning.AI · Stanford
May 2025Advanced Learning AlgorithmsDeepLearning.AI · Stanford
May 2025Supervised Machine Learning: Regression and ClassificationDeepLearning.AI · Stanford
OngoingHackathons — Yantra'25, Caterpillar, Bajaj Finserv, GrabContributor, OSCI Open Source
05

Contact

—

Got a problem worth automating?

Say hello

manavprakash91@gmail.com +91 95605 43184 github.com/Manav250305 linkedin.com/in/manavprakash25

Download my resume PDF, 2 pages, updated September 2026

Graduating May 2027 and open to full-time software and applied-ML positions, plus summer 2027 internships. Backend, data platform, or anything document-heavy is where I'm most useful.

Mostly LLM extraction pipelines, RAG systems, and full-stack prototypes that need to work in front of real stakeholders rather than in a notebook.

Getting models small enough for edge hardware, keeping LLM output structured and verifiable, or testing AI pipelines properly. Send a message either way.