# Sreenitya Thatikunta — full site text > Sreenitya Thatikunta is a Computer Science undergraduate at IIT Jodhpur who builds at the intersection of AI, machine learning and software engineering — diffusion and vision-language models, forecasting pipelines, and full-stack web products. Every page of https://www.sreenitya.in as markdown, in one file. Generated from the same source as the HTML, so it is never out of date. Structured JSON for the same content is at https://www.sreenitya.in/api. --- # Sreenitya Thatikunta > Sreenitya Thatikunta is a Computer Science undergraduate at IIT Jodhpur who builds at the intersection of AI, machine learning and software engineering — diffusion and vision-language models, forecasting pipelines, and full-stack web products. Canonical HTML: https://www.sreenitya.in/ ## About me Hello, I'm Sreenitya! I'm a Computer Science undergraduate at IIT Jodhpur who loves building things at the intersection of AI, machine learning, and software engineering. I enjoy taking ideas from “this would be cool” to something people can actually use, and I'm always excited to explore new technologies if they help solve the problem better. I like creating products that are simple, fast, and thoughtfully designed, with as much attention to the user experience as the code behind it. Whether it's an AI experiment, a web application, or a side project that started out of curiosity, I enjoy building things that are both technically interesting and genuinely fun to use. Outside of tech, I'm probably watching Formula 1 or cricket, hunting for a good coffee spot, or pretending I'll watch just one episode before inevitably finishing half the season. ## Skills ### Programming Languages - Python - C++ - C - SQL - JavaScript - TypeScript - HTML - CSS - Bash/Shell Scripting ### AI & Machine Learning - PyTorch - Scikit-Learn - OpenCV - Deep Learning - Transfer Learning - Time Series Models - LLMs - VLMs - Agentic AI ### Data Science - NumPy - Pandas - Matplotlib - Plotly - Optuna ### Web & Back end - React - Node.js - Express - FastAPI - MongoDB - Supabase - Redis ### Cloud & Data Platforms - Databricks - Snowflake ### Developer Tools - Git - GitHub - Linux - IntelliJ IDEA - VS Code - Postman - Streamlit - Weights & Biases - LaTeX ## Experience ### JPMorgan Chase & Co. — Summer Analyst (Jun 2026 - Jul 2026) - Built production forecasting pipelines using Prophet on 6+ years of banking transaction data. - Modelled trend, seasonality, holidays, and demand spikes for better capacity planning. - Delivered 92% forecasting accuracy during Amazon Prime business events. ### Pulsegen Pvt Ltd — Full Stack Developer Intern (May 2025 - Jul 2025) - Architected an AI-powered field management system with frontend, backend, and database design. - Integrated Mixpanel and ClickUp APIs to build analytics infrastructure and automated task workflows. - Built a platform-wide data export pipeline for application data and reporting portability. ## Projects ### Manus Fine-tuned PixArt-diffusion with LoRA on 3,000+ hand images and LLaVA captions. Reduced FID by 13.5% via Optuna tuning, benchmarked variants with FID and T2I-CompBench, tracked experiments in W&B, and deployed a Gradio demo. - [github](https://github.com/SreenityaThatikunta/PixArt-Alpha-LoRA) ### Visus Built a CLIP-guided cross-modal segmentation model for text-conditioned facial occlusion localization. Used synthetic training data, prompt augmentation, and Dice loss optimization to reach 94.4% mIoU on CelebA and 94.1% on LFW. - [report](https://www.sreenitya.in/Face_Segmentation_Report.pdf) ### Atlas Built a ResNet-50 image retrieval pipeline for CIFAR-10 with LDA reduction and classifier evaluation. Delivered 91%+ accuracy and deployed the system in Streamlit. - [github](https://github.com/SreenityaThatikunta/Image-Retrieval) ### Mini projects - terria: daily geography game — Like Wordle but with places — a daily geography guessing game. - lumière wallpapers — An Unsplash-backed wallpaper app for Android. - workout tracker — Built because no app had exactly what I wanted. - photobooth — Strike a pose, save the moment — a browser photobooth. - meme search extension — Indexed chaos, searchable — a browser extension for meme search. - stock market agent — Watches Indian markets so you don't have to. ## Contact Email sreenityathatikunta@gmail.com, or use the contact form on the homepage. - [LinkedIn](https://www.linkedin.com/in/sreenitya-thatikunta-b16a8a287/) - [GitHub](https://github.com/SreenityaThatikunta) - [Resume (PDF)](https://www.sreenitya.in/SreenityaT_Resume.pdf) - [Contact page](https://www.sreenitya.in/contact) ## For AI agents and developers This site publishes a read-only JSON API, an OpenAPI 3.0.3 specification, markdown variants of every page and an llms.txt index, so agents do not have to scrape the interface. - [API documentation](https://www.sreenitya.in/docs) - [Developer portal](https://www.sreenitya.in/developers) - [API index](https://www.sreenitya.in/api) - [OpenAPI specification](https://www.sreenitya.in/openapi.json) - [llms.txt](https://www.sreenitya.in/llms.txt) --- # About Sreenitya Thatikunta > Background, education and focus areas of Sreenitya Thatikunta, a Computer Science undergraduate at IIT Jodhpur. Canonical HTML: https://www.sreenitya.in/about Hello, I'm Sreenitya! I'm a Computer Science undergraduate at IIT Jodhpur who loves building things at the intersection of AI, machine learning, and software engineering. I enjoy taking ideas from “this would be cool” to something people can actually use, and I'm always excited to explore new technologies if they help solve the problem better. I like creating products that are simple, fast, and thoughtfully designed, with as much attention to the user experience as the code behind it. Whether it's an AI experiment, a web application, or a side project that started out of curiosity, I enjoy building things that are both technically interesting and genuinely fun to use. Outside of tech, I'm probably watching Formula 1 or cricket, hunting for a good coffee spot, or pretending I'll watch just one episode before inevitably finishing half the season. ## Education Computer Science Undergraduate at the Indian Institute of Technology Jodhpur. Coursework and project work centre on machine learning, computer vision and systems, alongside full-stack product engineering. ## What I work on - Artificial Intelligence - Machine Learning - Deep Learning - Computer Vision - Diffusion Models - Large Language Models - Agentic AI - Full Stack Web Development - Time Series Forecasting ## Elsewhere - [LinkedIn](https://www.linkedin.com/in/sreenitya-thatikunta-b16a8a287/) - [GitHub](https://github.com/SreenityaThatikunta) - [Resume (PDF)](https://www.sreenitya.in/SreenityaT_Resume.pdf) - [Contact](https://www.sreenitya.in/contact) --- # Contact Sreenitya Thatikunta > How to reach Sreenitya Thatikunta — email, contact form, LinkedIn and GitHub, plus a machine-readable contact endpoint for agents. Canonical HTML: https://www.sreenitya.in/contact The fastest way to reach Sreenitya Thatikunta is email: sreenityathatikunta@gmail.com. Messages sent through the form on the homepage arrive at the same address. Expect a reply within a few days — this is a personal site run by one person, not a support desk with a service-level agreement. ## Direct contact - **Email**: sreenityathatikunta@gmail.com - **Contact form**: https://www.sreenitya.in/#contacts - **LinkedIn**: https://www.linkedin.com/in/sreenitya-thatikunta-b16a8a287/ - **GitHub**: https://github.com/SreenityaThatikunta - **Location**: India ## What to get in touch about - Internship, full-time and research opportunities in AI, machine learning or software engineering. - Collaboration on open-source projects, especially in computer vision, diffusion models and agentic tooling. - Questions about any project listed on this site, including the reports and repositories linked from it. - Speaking to students about applying to and studying computer science at IIT Jodhpur. - Corrections — if something on this site is out of date, say so and it will be fixed. ## For automated agents Machine-readable contact details are available at /api/contact as JSON. Please do not use the address on this page for bulk or unsolicited commercial mail; the contact form and the mailbox are both read by a human. --- # Privacy Policy — Sreenitya Thatikunta > What https://www.sreenitya.in collects, why, which third parties are involved, and how to have your data removed. Canonical HTML: https://www.sreenitya.in/privacy This is the privacy policy for https://www.sreenitya.in, the personal portfolio of Sreenitya Thatikunta. It explains, in plain terms, what the site collects, why, and what it does not do. The site sells nothing, has no user accounts, and sets no advertising cookies. ## What is collected - Analytics: the site uses Google Analytics to count page views and understand which pages are read. Google Analytics records information such as your approximate location, browser, device type and the pages you visit. IP addresses are handled by Google under its own policies. - Contact form: if you submit the form on the homepage, the name, email address and message you type are sent to Formspree, which forwards them by email. That data is used only to reply to you. - Server logs: the site is hosted on Vercel, which keeps standard request logs (IP address, user agent, requested path) for operational and security purposes. - JSON API: requests to /api endpoints are subject to the same Vercel request logging. The API is read-only and unauthenticated, so no credentials are collected. ## What is not collected - No accounts, passwords or payment details — none of these exist on this site. - No advertising or cross-site tracking cookies. - No sale or sharing of personal data with data brokers. ## Third parties - Vercel — hosting and content delivery. - Google Analytics — aggregate usage measurement. - Formspree — delivery of contact-form submissions. - Google Fonts — web font delivery. ## Your choices You can block analytics with any content blocker; the site works perfectly without it. To have a contact-form message and the email it generated deleted, write to sreenityathatikunta@gmail.com and it will be removed. This policy may be updated; the current version is always the one published at https://www.sreenitya.in/privacy. --- # Sreenitya Thatikunta Developer Portal > Public, read-only JSON API, OpenAPI 3.0.3 specification, CLI and agent resources for the Sreenitya Thatikunta portfolio. Canonical HTML: https://www.sreenitya.in/developers Sreenitya Thatikunta's portfolio publishes a small, public, read-only JSON API so that AI agents, scripts and other developers can read the same information the website shows without scraping a JavaScript single-page app. Everything below is free to use, needs no API key, and is served over HTTPS with permissive CORS. ## Quickstart ``` curl -s https://www.sreenitya.in/api/profile curl -s "https://www.sreenitya.in/api/projects?kind=research" curl -s https://www.sreenitya.in/api/projects/manus ``` ## Authentication None. There are no API keys, tokens or sign-up. Every endpoint is public and read-only, so there is nothing to authorise. Please keep request volume reasonable — the whole dataset is a few kilobytes and responses are cached at the edge for an hour. ## Endpoints - **GET /api**: Index of every endpoint, with links to the spec. - **GET /api/health**: Liveness check. - **GET /api/profile**: Identity, headline, links, contact. - **GET /api/skills**: Skills grouped by category. - **GET /api/experience**: Roles with dates and highlights. - **GET /api/projects**: Projects; filter with ?kind= or ?tag=. - **GET /api/projects/{slug}**: A single project by slug. - **GET /api/contact**: Machine-readable contact channels. - **GET /api/openapi.json**: The OpenAPI 3.0.3 specification. ## Errors Every error is JSON, never HTML. The shape is stable: an error object with code, message, hint, documentation\_url and status. ``` { "error": { "code": "project_not_found", "message": "No project exists with slug 'nope'.", "hint": "List valid slugs with GET /api/projects.", "documentation_url": "https://www.sreenitya.in/docs", "status": 404 } } ``` ## Command line A dependency-free CLI wraps the same endpoints, so you can script against the site without writing an HTTP client. ``` npx sreenitya profile npx sreenitya projects --kind research npx sreenitya project manus --json ``` ## Agent resources - [OpenAPI 3.0.3 specification](https://www.sreenitya.in/openapi.json) - [API reference](https://www.sreenitya.in/docs) - [llms.txt](https://www.sreenitya.in/llms.txt) - [llms-full.txt (whole site as text)](https://www.sreenitya.in/llms-full.txt) - [Agent instructions](https://www.sreenitya.in/agent-instructions.md) - [Sitemap](https://www.sreenitya.in/sitemap.xml) ## Sandbox The production API is the sandbox: it is read-only, so no request can change anything. Use GET /api/health as a harmless smoke test when wiring up a client. --- # Sreenitya Thatikunta Portfolio API Documentation > Endpoint-by-endpoint reference for the Sreenitya Thatikunta portfolio API, including errors, filters and markdown content negotiation. Canonical HTML: https://www.sreenitya.in/docs Reference for the Sreenitya Thatikunta portfolio API — a public, read-only, unauthenticated JSON API over the content of https://www.sreenitya.in. Base URL: https://www.sreenitya.in/api. All responses are UTF-8 JSON with Access-Control-Allow-Origin: \*. The machine-readable contract is the OpenAPI 3.0.3 document at https://www.sreenitya.in/openapi.json. ## Conventions - Only GET, HEAD and OPTIONS are accepted. Any other method returns 405 with a JSON error. - Every successful response is an object; collections are returned under a named key with a count alongside. - Errors are JSON with error.code, error.message, error.hint, error.documentation\_url and error.status. - Responses are cached at the edge (Cache-Control: public, max-age=0, s-maxage=3600). ## GET /api/profile Returns the site owner's identity: name, headline, job title, affiliation, description, contact email, resume URL and social profiles. ``` curl -s https://www.sreenitya.in/api/profile ``` ## GET /api/skills Returns categories, each with id, title and a skills array of strings. ## GET /api/experience Returns experience, an array of roles with id, company, role, startDate, endDate (ISO year-month) and highlights. ## GET /api/projects Returns projects. Optional query parameters: kind (research or mini) and tag (case-insensitive exact tag match). An unknown kind returns 400 with invalid\_parameter. ``` curl -s "https://www.sreenitya.in/api/projects?tag=Diffusion" ``` ## GET /api/projects/{slug} Returns a single project. Unknown slugs return 404 with project\_not\_found and a hint pointing at the collection endpoint. ## GET /api/contact Returns the public contact channels: email, contact form URL and social profiles. ## Markdown content negotiation Every HTML page on this site also exists as markdown. Send Accept: text/markdown and you receive the markdown variant; the response carries Vary: Accept, Accept-Encoding. You can also request the markdown file directly by appending .md to the path. ``` curl -sH 'Accept: text/markdown' https://www.sreenitya.in/ curl -s https://www.sreenitya.in/about.md ``` ## Rate limits No hard limit is enforced. The dataset is static and edge-cached; please stay under a few requests per second and cache what you fetch.