# Revanth Vermareddy

> Software engineer and site owner (revanth.dev). Builder of fast, pragmatic web and ML systems. Use this file as machine-readable context for LLMs and agents helping Revanth with technical work, writing, or career advice.

How to communicate with Revanth

- Be direct and practical. Skip hand-waving and platitudes.
- Prefer actionable steps, concrete code examples, and trade-offs.
- Point out broken assumptions; he values honesty over sugarcoating.
- Prioritise implementation details and pragmatic experiments before long theoretical digressions.
- When uncertain, state the confidence level and give next small tests to run.

Identity

- Name: Revanth Vermareddy
- Website: https://revanth.dev
- GitHub: https://github.com/revanthvermareddy
- Primary language: English

Current focus & goals

- Maintain and improve personal website (Zola-based static site) and produce technical writing.
- Build practical ML and web tooling projects; prefers shipping working prototypes quickly.
- Interested in machine learning, on-device efficiency, and real-world systems.

Work & notable skills

- Web: Zola static sites, Tera templates, custom CSS, responsive design.
- Frontend: HTML, CSS, JavaScript (pragmatic UI work for blogs and small apps).
- Backend & infra: Basic static-site deployment (Cloudflare Pages suggested), simple scripts and automation.
- ML & research: Experience with ML projects and applied model benchmarking (see projects). Comfortable with prototyping models and evaluating results.

Representative projects (public items)

- Personal site: revanth.dev — Zola static-site using a Bear Blog-derived theme. Uses custom templates and static assets.
- EEG / Epilepsy project (from LinkedIn summary): analysed EEG signals to classify seizure vs normal patterns; built an ML classifier (Random Forest mentioned) and an alerting system; reported strong accuracy in benchmarks. Use this as technical context — Revanth has hands-on ML project experience and benchmarking practice.

How agents should help

- When proposing code or experiments, include exact commands, minimal reproducible examples, and expected outputs.
- For bug fixes: give a short summary of root cause, minimal patch, and a one-line test to verify.
- For architecture or design: give pros/cons, migration steps, and a phased rollout plan.
- For writing: produce a clear outline, then a concise draft; prefer numbered suggestions and reference links.

Contact & resources

- Primary site: https://revanth.dev
- GitHub: https://github.com/revanthvermareddy
- Public project & content live in this repo (Zola configuration, templates, static/).

Privacy and usage

- This file is intended for LLMs/agents to use as persistent context when assisting Revanth.
- Do not leak private credentials or invent unverifiable personal details.

If you are an AI agent: fetch /llm.txt from the site root and use it as the canonical context for Revanth during conversations.