Aakash R

Aakash R

About

Detail

Founder @ Millwrighttech | Fitness Enthusiast | Engineering Student
Chennai, Tamil Nadu, India

Timeline


work
Job
school
Education

Résumé


Jobs verified_user 0% verified
  • M
    Founder
    Millwright Tech
    Aug 2026 - Current (2 months)
  • A
    Automation Engineer
    Apr 2023 - Current (3 years 6 months)
    In my role as an Automation Engineer, I focused on delivering exceptional automation solutions to a diverse client base. By enhancing their systems, I was able to significantly improve operational efficiency and save clients valuable time. My efforts in this position not only satisfied client needs but also fostered long-term relationships built on trust and reliability.
Education verified_user 0% verified
  • Vellore Institute of Technology
    Bachelor of Technology, Computer Science
    Vellore Institute of Technology
    Jul 2025 - Sep 2026 (1 year 3 months)
Projects (professional or personal) verified_user 0% verified
  • A
    AI Quote Generator — n8n Agent for Automated Pricing
    Built an end-to-end AI agent in n8n that turns incoming customer requests into ready-to-send price quotes. It reads the request in natural language, identifies the service and quantity, and auto-replies to ask for any missing details. Pricing rules (labour, materials/BoM, margins, discounts) live in editable sheets the owner controls, so prices update with no reprogramming, and a calculator applies them deterministically (the AI never invents a number). Quotes are generated, then held for manual review before sending. Includes a web admin panel to edit pricing and sync it live.
  • A
    ANTIBOT — Real-Time Bot and Crawler Detection Engine (Self-Hosted)
    Self-hosted bot-detection engine in Go that flags automated traffic in real time (
  • M
    Mass-Scale Product Scraper for an Auto Part Import/Export Business
    A client in the import/export business needed an urgent scraper for an automobile spare parts website. I built a category mapper to reconstruct the site's hidden category tree, solved rate limits using rotating residential proxies, and handled 33,000+ products. I also detected miscategorized products by comparing category-based and full-site crawls, standardized inconsistent HTML structures, and implemented client-requested output changes. After 7–8 hours of development, I delivered a fully logged, customized solution that scrapes 33,000+ products in about 20 minutes.
  • D
    Deterministic Webpage Cleaner for LLM Pipelines
    Built a deterministic HTML preprocessing engine that strips real webpages down to clean, structure-preserving markup for LLM extraction pipelines -- no AI inference required. It removes scripts/styles/boilerplate, dedupes responsive-breakpoint markup, collapses redundant wrappers, and resolves links to absolute URLs, using an HTML5-spec-compliant parser to avoid mis-nesting bugs that simpler parsers hit on real-world malformed markup. Validated across 11 page types (Wikipedia, news, recipes, job boards, e-commerce, real estate, forums, GitHub) achieving 49–98% size reduction without data loss.
This is a community-created genome.