Hi, my name is

Sahil Mangotra.

Software Engineer → AI

I’m a passionate web app developer with a strong focus on modern web technologies to create websites that look great, feel intuitive, and perform flawlessly. Previously a software engineer, I’m now pursuing a master’s in Computer Science with a concentration in AI.

Sahil profile image

About Me

Sahil profile picture
I’m a seasoned software developer with a strong foundation in computer science and mathematics, passionate about web development and innovative tech solutions. I previously served as a Lead Software Engineer at uTrade Solutions in Mohali, India, where I led projects in algorithmic trading, optimized database operations, and guided teams in microservice architecture transitions to enhance trading performance and system efficiency. Currently, I’m advancing my expertise through a master’s in Computer Science with a concentration in AI. My research focuses on integrating AI into adaptive learning platforms, expanding my skills at the cutting edge of technology. Whether I’m mentoring junior developers, optimizing tech solutions, or exploring new AI applications, I bring a comprehensive skill set and a collaborative spirit to every endeavor. Here are a few technologies I've been working with recently:
  • Golang
  • Python
  • Postgres SQL
  • Kafka
  • React
  • PyTorch

Experience

Graduate Research Assistant - Kennesaw State University
August 2024 - Present
  • Engineered a production-grade RAG pipeline using LangChain and Llama-based models to dynamically synthesize curricula from complex technical documentation, enhancing adaptive learning delivery.

  • Designed a hybrid network simulator featuring autonomous AI agents for natural language topology generation and automated log analysis to translate system metrics into human-readable insights

  • Optimized the NSF-funded “QUINTET” agents using Chain-of-Thought and Sandwich prompting strategies, driving a 111% increase in Faithfulness (0.38 to 0.80) and achieving a 0.735 overall RAGAS validation score.

  • Architected a Reinforcement Learning from Human Feedback (RLHF) pipeline using Direct Preference Optimization (DPO), leveraging student preference datasets to align agent outputs for educational accuracy.

Lead Software Engineer - uTrade Solutions
Sep 2023 - July 2024
  • Led a cross-functional engineering team to plan and execute a major architectural pivot, migrating a legacy monolith to a scalable microservices architecture utilizing Golang and Kafka.

  • Achieved a 40% enhancement in the turnaround time for development and deployment cycles through this strategic microservices migration.

  • Developed a custom market data compression algorithm utilizing low-level C structs and smart packet concatenation, drastically reducing payload size and optimizing data reception speed by 30%.

Senior Software Engineer
Sep 2020 - Sep 2023
  • Developed a highly scalable algo-trading platform backend using Golang and Python, successfully handling 500+ concurrent custom trading algorithms and over 10,000 retail traders within the first quarter of launch. Demonstrating experience with high-volume, real-time data processing.

  • Designed high-throughput real-time data pipelines, processing hundreds of messages per second by implementing TCP sockets for internal microservices and WebSockets with custom compression for end-clients.

  • Optimized PostgreSQL databases handling millions of rows by implementing advanced indexing, sharding, and decoupling read/write replicas to support millisecond-level financial market data writes.

  • Collaborated independently with Exchange end-adapter teams to ensure seamless, low-latency market data integration.

Software Engineer
June 2010 - Sep 2020
  • Developed a new scalable REST-based backend using Django, functioning as a Platform-as-a-Service (PaaS) to allow multiple external B2B applications to connect seamlessly.

  • Developed a lightweight version of the core trading platform and built a cross-platform mobile application (iOS/Android) using the Ionic framework, driving new B2B client onboarding.

Education

Expected Graduation: 2026
Master of Science in Computer Science (AI Concentration)
Kennesaw State University
I am currently pursuing my Master of Science in Computer Science at Kennesaw State University, with a concentration in Artificial Intelligence. My graduate journey culminates in my thesis, QUINTET, a novel educational platform designed to democratize access to Quantum Computing. Recognizing that traditional tools are often fragmented and math-heavy, I developed a ‘zero-setup’ environment that combines gamified intuition, a hybrid network simulator, and RAG-based AI agents to guide students through complex concepts. This research has allowed me to merge my background in software engineering with advanced AI, building systems that are not just technically robust but pedagogically impactful.
2015 - 2019
Bachelor of Engineering in Computer Science
Chitkara University
GPA: 3.83 out of 4.0

I was part of the University Coding Academy. I passed a rigorous coding exam and interview, and was selected among a few students from the entire university. In uCA, we were taught complex subjects like advanced algorithms and advanced operating systems by external faculty associated with industry leaders like Infosys and Google.

Extracurricular Activities:

  • Participated in various coding competitions and hackathons.

Projects

QUINTET - AI Powered Quantum Education Toolkit
Python AI Quantum Cryptography LangChain Graduate Thesis
QUINTET - AI Powered Quantum Education Toolkit
This research introduces QUINTET, a novel containerized educational platform that eliminates this fragmentation by embedding interactive gamified modules and a customized quantum network simulation engine directly into a unified notebook environment. To provide real-time, context-aware support, the system integrates a dual-agent Artificial Intelligence framework driven by an active Retrieval-Augmented Generation pipeline, alongside a dynamic curriculum generator that personalizes learning paths based on a student's prior knowledge and time constraints.
QGrasp - Mobile app for quantum computing education
ReactNative Expo Computer Vision Quantum Circuts Graduate Thesis
QGrasp - Mobile app for quantum computing education
QuantumGrasp is a mobile app for quantum computing education. Students build circuits from physical playmats or on-screen controls, run on-device simulation, and see probabilities and circuit diagrams update in real time. The longer-term vision is a tactile, AR-forward experience (markers, 3D Bloch visualization, spatial measurement). The current build focuses on the learn → simulate → visualize loop on React Native / Expo.
Quantum Simulator
Python AI Quantum Cryptography LangChain Graduate Thesis
Quantum Simulator
QNS aims at facilitating Open-Ended Inquiry. Unlike traditional game-winning and game-losing conditions, QNS presents an open-ended scenario where students get to build their own network infrastructures and then watch their own designed functionality failing due to these modes, for instance, eavesdropping detection or decoherence. They develop from ‘sitting around and consuming knowledge’ to ‘actively building an understanding’ on their own.
Financial Knowledge Graph
Python Neo4j BERT FastAPI WebSockets
Financial Knowledge Graph
Constructed a large-scale Financial Knowledge Graph (FKG) by processing 350k+ news articles using a BERT-based relation extraction model (SMARTe) and spaCy to resolve over 200k entities into Neo4j. Engineered a custom "Ripple Effect" propagation algorithm to quantify and simulate systemic risk transmission across corporate networks in real-time.
Edge AI Optimization
Computer Vision YOLOv11 OpenVINO Edge AI Python
Edge AI Optimization
Fine-tuned YOLOv11n for object detection (tomato ripeness), achieving 95.5% [email protected] on a custom composite dataset. Optimized the model for resource-constrained edge devices using OpenVINO INT8 quantization, achieving a 4.3x inference speedup (6.5ms) with negligible accuracy loss.
DayCanvas
Generative AI Claude 3.5 Imagen 3 NLP Django
DayCanvas
An AI-powered journal visualizer that transforms written emotions into visual narratives. The multi-model pipeline integrates Claude 3.5 for emotional pattern extraction and Google's Imagen 3 for generating visual metaphors, creating a parallel universe that reflects the user's daily journey.

Get in Touch

My inbox is always open. Whether you have a question or just want to say hi, I’ll try my best to get back to you!