Amogh Agastya

Focused on building reliable, scalable, and intelligent agentic solutions using frontier LLMs and orchestration frameworks.

Architected and shipped two production agents for a global marketing
analytics platform: Influencer Discovery and Content Strategy, automating
research and campaign planning at scale.

Built modular, adaptive agents using Google ADK, applying planner–executor
patterns,

Led the development of multiple AI systems from architecture to deployment — specializing in agent workflows, advanced RAG pipelines, and intelligent API design.

Highlights:
🚀 Founding engineer behind Natasha, Builder.ai’s flagship AI product — built and scaled NLP services using LLMs and agentic design patterns.

🎥 Designed the Sizzle Reel Generation API, an autonomous agent workflow that genera

Spearheaded large-scale conversational AI systems for enterprise clients — focused on customer experience, automation, and user experience.

Highlights:
📈 Scaled chatbots to 100K+ monthly sessions, cutting support load by 85%+ via intelligent routing and NLP tuning.

🧩 Designed and deployed mega-agent architecture with 2000+ intents, improving accuracy and response quality across platforms.

🌐 Del

Built high-converting chatbots for SMBs across verticals — optimized flows for lead gen, appointment booking, and re-engagement.

Highlights:
💰 Drove $100K+ in revenue via sales-oriented bots for ecommerce and service-based businesses.

🤖 Delivered full-stack solutions using Manychat, Twilio, and Messenger APIs — from flow design to deployment and analytics.

Contributed to backend development and API testing within CI/CD pipelines, following Agile and Behaviour Driven Development (BDD) practices.

Highlights:
🔧 Built and validated REST APIs using Java Spring MVC and Python Robot Framework.

⚙️ Automated test workflows to enhance reliability and reduce manual QA effort.

📦 Gained hands-on experience with Docker, Git, MySQL, and backend release cycles.

Applied Natural Language Processing (NLP) techniques to extract insights from job descriptions and resumes using modern language models and data mining.

Highlights:
🧾 Built a NER (Named Entity Recognition) model for skill extraction, by finetuning BERT using spaCy, deployed on Google Cloud GPU.

📊 Created a prototype to semantically match candidates to job descriptions using word embeddings.

🌐 D

Pursued a Bachelor of Engineering in Computer Science from B. M. S. College of Engineering from 2016 to 2020.