SPTK

Software Engineer · Systems & AI

Sai Pavan Teja
Kunche

Building scalable software systems from distributed task queues and custom search engines to LLM-powered RAG pipelines and aerospace-grade Digital Twin simulators. Full-Stack Builder. Systems & AI-obsessed.

Distributed Systems Algorithmic Design Information Retrieval RAG · LLM · Vector Retrieval Digital Twin Architecture
View Projects Connect
01 About

The Mind Behind
the Architecture

Sai Pavan Teja Kunche
0
Years Experience
0
Systems Engineered
0
Certifications
0
Test Cycles Led

Software Engineer & Full-Stack Builder

I'm a Software Engineer specializing in Python, data engineering, and generative AI application development. My work focuses on building resilient ETL and validation pipelines, integrating LLMs via production APIs, and deploying high-accuracy RAG workflows using vector embeddings and similarity search on AWS.

Grounded in core computer science fundamentals, I build scalable systems from scratch. That spans engineering fault-tolerant distributed task queues with Python sockets and multiprocessing, building inverted-index search engines with min-heap retrieval, and developing production digital twin simulators with zero-dependency MAVLink v2 bridges.

Beyond the backend, I deliver responsive full-stack applications and automated enterprise workflows. I focus on bridging complex algorithmic logic with clean user interfaces, ensuring high concurrency, strict type safety, and verifiable system outputs.

Systems Engineering Concurrency Algorithms RAG Pipelines LLM Integration Information Retrieval TypeScript / React Python / FastAPI Digital Twin Architecture
02 Experience

Where I've
Built Things

Associate Data Analyst
Camp Systems International
Hyderabad, India · June 2021 – June 2023
  • Designed automated data mapping & validation workflows for legacy aviation records, improving accuracy by +15%
  • Built SQL-driven dashboards monitoring fleet maintenance health for data-backed operational decisions
  • Reduced data processing turnaround by 20% through structured validation and transformation logic
  • Authored standardization & compliance documentation accelerating onboarding across technical teams
Research Supervisor
Abyom Space Tech
Gujarat, India · March 2021 – May 2021
  • Led a team of junior engineers in the R&D roadmap for subsystem simulations using AWS cloud resources
  • Orchestrated static fire testing campaigns and oversaw DAQ system setups with MuJoCo integration
  • Validated performance across 20+ successful test cycles with precision measurement protocols
03 Projects

Engineered Systems &
AI Pipelines

06
Flagship Project
Copter Studios
Aerospace-Grade Flight Simulator · v12 · Hardware-in-the-Loop

A production-grade Copter Studios with full-fidelity physics, a from-scratch MAVLink v2 HIL bridge that connects to real PX4/ArduPilot flight controllers, and an AI sim operator powered by Gemini. Built with the same standards used in professional aerospace simulation then deployed in a browser.

Physics Core
RK4 integration · BET thrust · Quaternion attitude · Dryden turbulence · ISA atmosphere
HIL Protocol
MAVLink v2 encoded from scratch (zero deps) · HIL_SENSOR · PX4 & ArduPilot compatible
RL Agent
PPO Actor-Critic · TensorFlow.js · Web Worker concurrent benchmarking
AI Operator
Aether Gemini-powered sim controller · NL→config parsing · Crash forensics
MAVLink v2 HIL Digital Twin RK4 + BET Physics TypeScript / React TensorFlow.js PPO / RL Gemini API PX4 / ArduPilot Web Workers
Launch Simulator copterstudios.com
System Architecture
PX4 / ArduPilot SITL ⟷ MAVLink v2 UDP Bridge
HIL_ACTUATOR_CONTROLS ↓ · HIL_SENSOR ↑ · HIL_GPS ↑
PhysicsEngine v12 → DroneState ×62.5Hz
RK4 · BET · Quaternion · Dryden · ISA · Ground Effect
RLAgent (Actor-Critic) → Episode Worker
TF.js · PPO · User model hot-swap · Concurrent execution
Aether AI Operator → SimulationParser
Gemini Flash · NL→PhysicsConfig · Approval Dialog
Why This Stands Out

MAVLink v2 CRC-16/MCRF4XX implemented from scratch with zero npm dependencies. Connects this browser-based simulator directly to a real Pixhawk flight controller or PX4 SITL the same architecture used in professional aerospace simulation toolchains.

01
MiniIndex Search Engine
Algorithmic Information Retrieval

Built an inverted-index search engine from scratch featuring a custom tokenizer, BM25 ranking algorithm, and min-heap-based top-K retrieval to avoid full-set sorting. Complete with boolean query parsing and Levenshtein-distance typos.

PythonInverted IndexBM25HeapsAlgorithms
View Source
02
TinyQueue
Distributed Task Queue

Engineered a fault-tolerant job queue using a socket-based client/server protocol and concurrent multi-process workers. Implemented at-least-once delivery via visibility timeouts and dead-letter queues after max retries.

PythonSocketsMultiprocessingDistributed Systems
View Source
03
TerrierHelper AI
Institutional RAG Assistant

Engineered a client-side RAG system ingesting PDFs to deliver policy-accurate answers via LLM embeddings. Strict prompt guardrails eliminate hallucinations with real-time token streaming and cosine-similarity searches.

Gemini APIReact / TSRAGVector Retrieval
View Source
04
University Smart Outlook
LLM-Powered Email Automation

AI tool analyzing incoming emails and auto-drafting context-aware responses from a secured knowledge base. Human-in-the-loop approval with semantic analysis for consistent administrative tone at scale.

Gemini FlashPythonWorkflow AutomationNLP
View Source
04 Skills

Technical
Proficiency

Software & Systems Engineering
Data Structures & AlgorithmsADVANCED
Concurrent ProgrammingADVANCED
Distributed Systems (Sockets)PROFICIENT
Client-Server ArchitectureADVANCED
Information Retrieval (BM25)PROFICIENT
AI, ML & Simulation
RAG & Vector EmbeddingsADVANCED
LLM Integration (Gemini, Claude)ADVANCED
Reinforcement Learning (PPO)PROFICIENT
Physics Engine Design (RK4)ADVANCED
MAVLink v2 Protocol & HILADVANCED
Languages, Tools & Cloud
Python (AI, Backend & Algorithms)ADVANCED
TypeScript / JavaScriptADVANCED
React.js / Node.jsADVANCED
Google Cloud / Vertex AIPROFICIENT
Docker / AWS / GitPROFICIENT
05 Education

Academic
Foundation

🎓
Master of Science, Information Technology Management
St. Francis College, Brooklyn, NY
Aug 2023 – May 2026
🚀
Bachelor of Technology, Aerospace Engineering
Lovely Professional University, India
Aug 2017 – May 2021
06 Contact

Let's Build Something
Extraordinary

Actively seeking Software Engineering and Applied AI roles. Whether you're building intelligent autonomous systems, scalable infrastructure, or the next wave of LLM-powered applications let's connect.

✉saiptkunche@gmail.com ☎+1 551 208 4764 inLinkedIn ⌨GitHub