I'm Konchada Mahadev — A Btech graduate in Electronics and Communication Engineering and Specialist Programmer with practical experience delivering full‑stack Python services and Gen‑AI integrations. My work spans document intelligence platforms, automation pipelines, and developer-facing foundation services that make LLMs practical for production use. I enjoy turning research ideas into reliable, maintainable systems and mentoring peers on practical ML/AI integrations.
Core strengths: Python, Flask/FastAPI, SQL databases,AI,ML, Document parsing & retrieval, LLM orchestration, and API design.
Focus: Building scalable document ingestion & retrieval systems, pragmatic Gen‑AI features (summarization, retrieval-augmented generation), and developer tooling that improves team velocity.
Contact: mahadevsai12@gmail.com Phone no: +91 9346821030
Company: Bluerose Technologies and Private Limited
Client: Oppo & Oneplus Research and Development Center
Duration: July 2023 - Feb 2024
Company: Bluerose Technologies and Private Limited
Client: Oppo & Oneplus Research and Development Center
Duration: Feb 2024 - Sep 2024
Company: Infosys
Duration: Sep 2024 - Present
Python, GEN-AI, Vector Databases, Embeddings, RAG (Retrieval-Augmented Generation), Azure AI Search
PyCharm, SQL Workbench, Jupyter Notebook, VSCode, GIT
Flask & FastAPI, LangChain & LangGraph, Prompt Engineering, Machine Learning, Front-end (Streamlit / HTML & CSS)
Board: CBSE
Percentage: 78%
Group: MPC with Computer Science
Board: CBSE
Percentage: 81%
Branch: Electronics and Communication Engineering
Board: JNTUV
CGPA: 8.56
Foundation Platform – Reusable Document Intelligence APIs
Designed and developed reusable APIs for document ingestion, retrieval, summarization, and LLM invocation, forming the core foundation for multiple downstream projects.
Implemented asynchronous ingestion workflows with support for .pdf, .docx, .txt, etc., enabling background processing and job tracking through a status API (in-progress, completed, failed).
Built retrieval APIs to fetch documents based on user queries and session context, improving traceability and user experience.
Developed summarization using MapReduce techniques for both documents and LLM conversations, enhancing efficiency and scalability.
Built an AI-driven assistant to automate proposal creation by ingesting documents (.pdf, Box links, etc.) into a vector database for intelligent retrieval.
Developed pipeline to extract and structure proposal data (text, tables, images) and populate a standardized template using LLMs.
Integrated best practices documents into the LLM workflow to ensure proposals adhere to organizational guidelines and quality standards.
Implemented advanced document parsing and formatting to handle complex elements like tables and images seamlessly.
Delivered a scalable solution that reduced manual effort in proposal preparation and improved consistency and turnaround time.
Uses Google Gemini (configurable) to generate clean, documented Python code from natural-language prompts and returns both code and generation metadata.
Supports safe code execution in a sandboxed environment with captured stdout/stderr and execution success tracking.
Stores user sessions and generation history in a database (SQLite by default, MySQL optional) with analytics endpoints and a server-rendered history UI.
Includes developer tooling and setup scripts (init_db, setup_database, query_db, check_db) and a `config_template.py` for easy configuration.
Designed with security and observability: API key hashing for analytics, input validation, error logging, and endpoints for history and analytics (/generate, /execute, /history, /analytics).
Mig-150, Near 7 Temples, Babametta , Vizianagaram, AndhraPradesh, India, 535002
Phone+91 9346821030
Emailmahadevsai12@gmail.com
