AI/ML engineer Full stack web developer Hyderabad, India

PoojithaBheemreddy

I build AI systems that have to be right, and the web apps around them: computer vision, LLM workflows, and React and Next.js interfaces that let people check the answers.

psst. You're reading me in plain visible light. I look way cooler in and . Flip the switch ↗ Welcome to UV. This is where the hidden stuff glows. Go back to or try for heat vision. Heat vision on. Hot take: I'm very good at imaging. Cool down in or peek at .

$ whoami
poojitha · B.Tech CSE, GRIET · CGPA 9.72
$ papers --status
2 under review (IEEE . · AIP Conf. Proc.)
$ stack --web
React · Next.js · Node · Express · Tailwind
$
Portrait of Poojitha Bheemreddy
Profile

Models on one side. Interfaces on the other.

About me, in three lights

I'm Poojitha, a computer science engineer from Hyderabad. I train vision models, and I build the web apps around them.

  • Web developer: React, Next.js, Node, Express and Tailwind, from Solytics to a 3D product-visualization frontend at Wearcvnt.
  • B.Tech CSE at GRIET with a CGPA of 9.72, top 2% of 432.
  • Interned at Google (AI/ML), Accenture, ServiceNow and Solytics.

Before code there was maths. The habit of wanting a ground truth shows up in everything I build.

  • Asian International Mathematical Olympiad gold (national rank 3) and SEAMO silver (global rank 3).
  • Harvard HUIRSP scholar, studying AI ethics and governance. Kode with Klossy and Girls Who Code alum.
  • Led a 52-member Computer Society of India team and ran E-Summit for 400+ people.

What I can't leave alone: images that hide something. Cracks in microscopy, forged holograms, a second ink.

  • Two imaging papers under review, on nano cracks and fake certificates.
  • National finalist at CodeQuest 2026 with ClaimGuard AI, a human-in-the-loop claims tool.
  • WorldQuant IQC gold, hunting alphas in market data.

Computer science engineer from GRIET, Hyderabad, top 2% of 432 in the department. I train models with TensorFlow and OpenCV, then ship them behind React and Next.js front ends with monitoring and CI/CD.

9.72CGPA, B.Tech CSE
2Papers under review
3Olympiad medals
108CodeQuest national finalists of 2,100+
2022 – 2026B.Tech, Computer Science and EngineeringGokaraju Rangaraju Institute of Engineering and Technology9.72
2019 – 2021Intermediate, MPCSri Chaitanya Junior Kalasala, Telangana State Board99.2%
2018 – 2019Secondary School CertificateSri Chaitanya High School, Telangana State Board
Data Structures & AlgorithmsOOP (Java)DBMSOperating SystemsComputer NetworksSoftware EngineeringCloud ComputingMachine LearningAINLPFull Stack Web
Experience

Six roles across frontend, automation and ML.

Open any row for what I shipped.

  • Built scalable, responsive web apps in React.js, ES6+ JavaScript, HTML5, CSS3 and Tailwind CSS, with modular reusable components and tuned state management and rendering.
  • Integrated REST APIs and asynchronous data flows with Axios.
  • Completed an Application Performance Monitoring (APM) logger that tracks, logs and monitors application performance metrics in real time.
  • Debugged and profiled with browser developer tools; worked through testing, deployment and CI/CD.
React.jsJavaScript ES6+Tailwind CSSAxiosAPMCI/CD
  • Engineered a high-performance frontend in Next.js (SSR and CSR) and React.js with component-based architecture, tuning rendering and load times.
  • Built reusable UI modules with Tailwind CSS and managed state and props for scale.
  • Brought 3D assets and animation into the UI pipeline for real-time product visualization.
Next.jsReact.jsTailwind CSS3D assets
  • Designed and deployed Blue Prism RPA workflows with object-layer abstractions, process-layer orchestration and exception handling for enterprise logistics automation.
  • Built Java automation modules integrated with enterprise APIs and databases for end-to-end process execution.
  • Applied Generative AI concepts to automation pipelines: decision logic, data-driven triggers, throughput tuning.
Blue PrismJavaRPAGenerative AI
Certificate ↗
  • Built and automated workflows in Flow Designer with business-process logic and event-driven automation.
  • Configured platform modules: reports, forms and system administration (CSA-level concepts).
  • Designed and tested enterprise workflows with the Automated Test Framework (ATF) and explored Agentic AI for service management.
Flow DesignerATFAgentic AI
Certificate ↗
  • Trained TensorFlow models with supervised and unsupervised learning on real-world datasets.
  • Handled preprocessing, feature engineering and evaluation with cross-validation and regularization.
  • Built neural-network solutions for image classification and object detection.
TensorFlowNeural networksCross-validationObject detection
Certificate ↗
  • My first taste of research. As a young research student I worked under Ms. Martha Guerrero at the New York Academy of Sciences, on engineering materials for automotive energy applications.
  • It is where I learned to treat a result as something that has to be checked, a habit that now shapes how I build and test my own models.
Engineering materialsAutomotive energyResearch
Document ↗
Lab

Two imaging papers, running in your browser.

Both papers are about seeing what ordinary images hide. These demos are small, simplified versions you can poke at.

Under review · 2026

Nano crack detection using residual imaging

A nano-scale crack framework built on Edge Residual Imaging: advanced preprocessing, feature extraction and CNN-based segmentation for precise structural defect analysis.

The demo generates a noisy synthetic micrograph with a hidden crack, then walks it through a classical pipeline: smooth, subtract, threshold, clean up. Because the crack is known, the page scores itself.

  • Submitted to International Conference on Computational Intelligence in Data Science
  • Authors Bheemreddy Poojitha, R. Aruna Flarence, Mentapally Vaishnavi, Sudini Sanjana Reddy, Tumma Anusha

Illustration only. The paper's method adds CNN-based segmentation; this uses a classical filter chain on synthetic data.

detection224 × 224 · synthetic
0detected px
0precision
0recall
0F1 (±1 px)
SAMPLE INSTITUTE OF TECHNOLOGY Certificate of Achievement THIS SPECIMEN CERTIFIES THAT A. Student PROGRAMMEB.Tech, Computer Science GRADEFirst Class with Distinction ISSUED30 May 2024 SAMPLE INSTITUTE · SPECIMEN · SAMPLE INSTITUTE · SPECIMEN · SAMPLE INSTITUTE · SPECIMEN · SAMPLE INSTITUTE · SAMPLE INK MISMATCH: GRADE FIELD SPECIMEN · NOT A REAL CREDENTIAL
VISEverything looks genuine. The altered grade matches the other lines.
UVHologram ring, fluorescent fibres and microtext light up. Photocopies and prints lose them.
IRThe edited grade used a different ink and drops out. Flagged as a mismatch.
Under review · 2025

Fake certificate detection with multispectral imaging and AI

An AI-driven forgery detector that images documents in UV and IR to verify holograms, analyze microtext and extract spectral features. Designed for real-time authentication on web and mobile.

The specimen on the left reacts to the page's spectral switch. Change the light and the hidden security layers appear.

  • Submitted to AIP Conference Proceedings
  • Authors R. Aruna Flarence, Bheemreddy Poojitha, Tumma Anusha, Mentapally Vaishnavi, Sudini Sanjana Reddy
  • Stack CNN, Siamese networks, Fourier hologram analysis, OCR microtext, Flask

Illustration with a fictional specimen. It is not a real document or a working verifier.

Projects

Things I built end to end.

Hackathon · Apr 2026Top 108 of 2,100+

ClaimGuard AI

A healthcare automation tool for medical claims with strict human-in-the-loop logic, so automation speeds the work up without letting errors through. Built for the CodeQuest 2026 GenAI Hackathon and took me to the national finale.

GenAIHuman-in-the-loopHealthcareResponsible AI
Research · Jul 2025 – Jun 2026Run demo →

Nano Crack Detection

Extracts high-frequency structural defects from microscopy images with Edge Residual Imaging. Gaussian filtering, denoising, normalization, contrast enhancement, then Canny, Sobel and Laplacian edges feeding CNN and Vision Transformer segmentation.

ERICNN / ViTOpenCVMorphology
Research · Jan – May 2025See specimen →

Fake Certificate Detection

An authentication system on UV, IR and visible imaging. CNN and Siamese models classify forgeries and learn similarity, with Fourier-based hologram detection, OCR microtext extraction, and a Flask interface for real-time verification.

Siamese netsSpectral analysisOCRFlask
Dec 2023 – Jun 2024

Cloud Document Collaboration

A real-time collaborative document platform with concurrent editing, version control and multi-user sync. REST APIs handle storage, access control and sessions, while WebSockets and event-driven updates keep every device consistent.

WebSocketsREST APIsVersion controlAccess control
Skills

Drag the sphere. Filter the stack.

Everything here comes from work above. Pick a group on the right to light it up.

drag to rotate
drag · hover · pick a group →
Awards

Competitions, olympiads and ranks.

Leadership

Teams, events and programs.

Executive Lead, Computer Society of IndiaOct 2023 – 2026

Led a 52-member team running workshops, ideathons and resume-building sessions for 500+ participants, including speaker onboarding.

Vice President, Entrepreneurship CellDec 2022 – 2026

Planned and ran E-Summit with industry speakers and 400+ participants; built ties with startups, founders and experts.

Hyderabad Section Co-Lead, IEEEXtreme 20232023

Coordinated a global programming contest of 16,000+ participants across 69 countries and ran operations across Hyderabad colleges.

Deputy Class Representative, GRIET2022 – 2026

Coordinated academic communication between students and faculty.

Amazon ML Summer SchoolAug 2025 · selected participant

Invite-only program by Amazon Scientists covering supervised and unsupervised learning, deep networks, generative AI and LLMs, reinforcement learning, and causal inference.

Harvard Undergraduate International Relations Scholars Program2022 · scholar

Explored AI ethics, governance and global policy, plus socio-economic issues. Certificate ↗

Kode with Klossy Scholar2022

Global coding program on web and app development, built to empower women in technology. Certificate ↗

Girls Who Code, Summer Immersion2022

International leadership and global issues program on social impact and cross-cultural collaboration. Certificate ↗

Contact

Let's build something that has to be right.

Emailpoojithabheemreddy111@gmail.com
LinkedIn/in/poojitha-bheemreddyOpen profile ↗
Languages
TeluguMother tongue
EnglishC1–C2 · IELTS 8.0
HindiC1–C2
8.0IELTS overall · C1
8.0Listening
8.5Reading
8.0Writing
7.0Speaking