Praneeth Mangala

AI systems engineer focused on agentic AI, RAG pipelines, LLM orchestration, and scientific ML. Experienced in building scalable AI products, automation systems, and full-stack applications across research, fintech, and enterprise domains.

• Built agentic AI workflows for enterprise productivity automation and communication intelligence.
• Designed Meeting Transcript → Linear task automation using LLM-based structured extraction.
• Developed RAG-based memory systems for persistent contextual reasoning.

• Built React dashboards for real-time fleet analytics and business insights.
• Processed 100K+ daily records via optimized SQL pipelines, improving performance by 30%.
• Developed ML models for bottleneck detection, reducing system errors by 20%.

• Developing Physics-Informed Neural Networks for nonlinear PDE systems.
• Embedding physics constraints into neural networks for improved convergence and stability.

• Designed and developed a guided molecular communication hardware prototype achieving a Bit Error Rate (BER) below 5%.
• Successfully demonstrated the working prototype at the National Conference on Communications (NCC'25), IIT Delhi.
• Presented an IEEE research paper at ICMOCE'25 hosted by IIT Bhubaneswar.
• Filed a patent for the proposed molecular communication hardware system.

B.Tech Electronics and Communication Engineering | CGPA: 3.09/4

HSC | 98.6%

AI Co-Founder System

• Agentic AI system converting startup ideas into execution plans.
• Planner, Validator, Executor multi-agent architecture.
• Cognee AI + RAG memory for contextual reasoning.

App Graph Builder

• ReactFlow-based visual workflow builder for modular applications.
• Zustand state graph engine for node lifecycle management.
• Extensible system for agentic workflows.

StayEase PG System

• Full-stack PG management system with Firebase backend.
• Real-time tenant tracking, payments, and complaints.
• Role-based admin architecture for scalability.

PINN Oscillator Model

• PyTorch-based physics-informed neural network for ODE simulation.
• Integrated physics loss constraints for stable learning.

• Winner – RBI Global Hackathon HaRBInger 2026
• Global Finalist – Global Fintech Fest 2025
• Jury Award – HackIITK Cybersecurity Hackathon
• IEEE Conference Publication – ICMOCE 2025
• Presented at IIT Delhi NCC’25
• Filed Patent – Molecular Communication System
• Top 1% in JEE Mains among 1.2M candidates