Headline

R&D Engineer @ Jukshio · MS Purdue · Agentic AI (LangGraph & CrewAI) · FastAPI & MCP Tools

About

I basically work on building AI systems that actually *do stuff*, not just chat back fancy replies 😄 Most of my work is about making AI useful in real-world situations, like connecting LLMs to tools, APIs, and workflows. I spend a lot of time with things like LangGraph and CrewAI to build these stateful systems that can remember context and take proper actions. I also work a lot on the backend side with Python, FastAPI, and databases, handling all the plumbing so things don’t break when they scale. I did my Master’s from Purdue, but honestly most of my learning comes from building things myself. I’ve been working on projects like Kueri, which converts natural language into SQL, and Sleuth, which automates invoice processing using AI. At the end of the day, I just enjoy solving messy problems and figuring out how to make AI actually reliable and practical, not just cool demos 😄

Location

Hyderabad

AI stack

cursor

deepseek

langgraph

mcp

openai

huggingface

Work experience

Python AI Engineer at Jukshio, Jun 2024 - Present · 2 yrs 6 mos, Hyderabad, Telangana, India, Main Projects & Products • Led the design, development, and launch of three main AI products: MCP Studio, AB Studio, and Agentic Studio. • Built MCP Studio, allowing users to easily turn APIs into tools and connect them with AI models without writing complex code. • Created AB Studio to test AI agents, track performance metrics, compare different models, and fix errors. • Built Agentic Studio to create AI workflows with LangGraph and added tracking tools to monitor system performance. • Delivered end-to-end AI software by connecting backend systems, AI agent frameworks, and monitoring tools. Technical Contributions & Results • Built a multi-agent system using CrewAI to send user requests to the correct tools automatically. • Designed a main supervisor agent that divides difficult tasks among smaller, specialized helper agents. • Created step-by-step AI workflows in LangGraph to handle repetitive tasks and agent collaboration. • Built fast web services using FastAPI that connect to Google Vertex AI models and secure login systems. • Set up over 27 custom tools with clear rules to ensure safe and reliable execution. • Built a streaming system using FastMCP to send live data to image and text recognition tools. • Connected MCP tools to CrewAI across 15 points, improving tool success rates by 40%. • Designed clean PostgreSQL databases using SQLAlchemy and Alembic for smooth updates. • Improved API speed by caching common data, making responses 30% faster. • Wrote Python scripts to automate daily work, cutting manual effort by 40% and speeding up document tasks by 75%. • Packaged applications using Docker and Kubernetes for cloud hosting, raising server efficiency by 40%. • Updated older Python software to modern versions, keeping production systems stable and reliable. • Built machine learning pipelines with Transformers, Flask, and TensorFlow to run AI models locally.

Python Developer at Rubistone Technologies LLC, Feb 2023 - Aug 2023 · 7 mos, Chicago, Illinois, United States, • Built Python APIs and Flask/Django services, making communication between backend and frontend systems 50% faster. • Created a web dashboard for weather data that loaded in under 500 ms and worked on almost any device. • Organized large datasets with Apache Parquet, making data searches 40% faster and reducing storage costs by 25%. • Processed over 1 million records using PySpark on AWS, speeding up data processing by 35%. • Set up cloud file storage and automatic data sync using AWS S3, Boto3, and AWS Lambda. • Improved database queries and search speeds by 50%, reducing page load times. • Built automated systems with AI to process documents, reducing manual review time by 30–40%. • Designed an AI system to collect and organize over 500,000 news articles daily with 99.5% uptime. • Automated Azure cloud setup using Terraform scripts, cutting setup time by 40%.

Python Backend Developer at Jukshio, Jul 2020 - Aug 2021 · 1 yr 2 mos, Hyderabad, Telangana, India, • Developed and maintained Python applications using libraries such as NumPy, Django, Flask, UnitTest, and Py Test, ensuring code quality and reliability. • Designed, developed, and maintained automated test scripts using Python and the selected automation framework for web and/or mobile applications. • Integrated external systems and databases, including Oracle DB, for data retrieval and storage, enhancing application functionality. • Implemented database schemas, crafted complex SQL queries, and optimized database performance for efficient data management. • Conducted comprehensive unit testing using frameworks like UnitTest and PyTest, ensuring code correctness and reliability. • Utilized version control systems such as Git and hosted repositories on platforms like GitHub and Bitbucket for efficient code management and collaboration. • Collaborated with QA teams to identify and address software defects and actively participated in code reviews to maintain code quality standards. • Worked within an Agile development environment, participating in daily stand-ups, sprint planning, and retrospectives. • Engaged actively in the software development lifecycle, from requirements gathering and design to implementation and maintenance. • Contributed to continuous improvement initiatives, sharing knowledge, and exploring new technologies and best practices to enhance team productivity and project efficiency.

Education

Master of Science - MS, Computer Information Technology, Purdue University

Bachelor of Technology, Electrical, Electronics and Communications Engineering, Aditya College Of Engineering & Technology

Public links

github: https://github.com/Ramc26

medium: https://medium.com/@itsmeramc

rambuilds.in: https://rambuilds.in/