About Me

Professional Experience

I ship AI systems to production, then study why they work. My approach: research rigor meets deployment reality. From RAG systems serving thousands of users to multi-agent AI for power grids, I build technology that scales responsibly for infrastructure that lasts.

Shreya Savant

Current Role

Graduate Research Assistant focused on multi-agent AI for power grid optimization.

Concordia University VoltAge Doctoral Fellow

Research Experience

Academic positions focused on AI systems research and development.

Graduate Research Assistant

Concordia University • VoltAge Fellowship • Sep 2025–Present

Investigating coordination mechanisms for distributed AI agents in power grid management and climate modeling. Developing novel optimization algorithms for multi-agent decision-making under uncertainty. Collaborating with industry partners on real-world deployment of research prototypes.

Software Development Engineer

SpokenWeb (SSHRC/Concordia) • Jun 2024–Present

Architected automated data pipelines for cultural heritage metadata processing using Ruby, Python, and Apache Solr. Improved processing efficiency by 61% through algorithm optimization and parallel processing. Modernized infrastructure with Docker containerization and CI/CD, reducing deployment time by 80%. Presented technical architecture at Blacklight Summit 2025.

Research Team Lead

Innovation Lab • Jan 2024–Dec 2025

Led 4-person research team developing multilingual conversational AI system (6 languages) for cyber-violence prevention. Designed novel prompt-driven persona framework using Gemini LLMs with Redis-backed conversation state. Developed automated UI/UX evaluation methodology using GPT-4o, reducing testing cycles by 40%. System deployed to 2,000+ users.

Industry Experience

Production engineering roles with measurable business impact.

Applied ML Researcher

7Dish (Contract Research) • Jun 2023–Jan 2024

Researched and developed production RAG architecture using LangChain and GPT-4 for nutrition question-answering. Investigated BERT-based embedding models for semantic search, improving retrieval accuracy by 21% on Databricks. Designed real-time ETL pipeline for knowledge base enrichment. Deployed system on Azure serving 5,000+ monthly active users with 30% engagement increase.

Data Science Intern

Visive Inc. • Aug 2022–Dec 2022

Developed real-time image search engine using TensorFlow on AWS, reducing false positives by 16%. Built fraud detection pipelines using statistical modeling, achieving 92% accuracy in production. Engineered ETL pipelines processing 1M+ daily transactions with optimized database schemas.

Software Developer Intern

ProfCess • May 2021–Aug 2021

Developed ML-driven recommendation system using scikit-learn, improving content relevance by 40%. Enhanced frontend performance and mobile UX across core user flows.

Technical Skills

Core competencies across research, engineering, and deployment.

AI/ML Research

PyTorch, TensorFlow, Hugging Face Transformers, Weights & Biases, MLflow, TensorBoard, Jupyter, Google Colab

LLM Systems

GPT-4, Claude, Gemini, Llama, Mistral with fine-tuning, prompt engineering, RLHF. LangChain, LlamaIndex, LangGraph, RAG architectures

Vector & Search

Pinecone, Weaviate, Apache Solr, Blacklight. Semantic search, embedding optimization, knowledge retrieval systems

Programming

Python (expert), JavaScript/TypeScript (proficient), Go, SQL, Ruby. FastAPI, Flask, Django, React, Node.js, Ruby on Rails

Infrastructure

AWS (EC2, Lambda, S3, SageMaker), Azure AI, Google Cloud. Docker, Kubernetes, CI/CD (GitHub Actions), Apache Airflow

Data Engineering

Apache Spark, PySpark, Databricks, ETL pipelines, PostgreSQL, MongoDB, Redis. Real-time data processing and pipeline optimization

Research Methods

Experimental Design, Statistical Analysis, A/B Testing, Model Evaluation & Benchmarking, Ablation Studies, Scientific Writing

Collaboration

Cross-functional team leadership, research mentorship, technical documentation, conference presentations, stakeholder communication

Achievements & Recognition

Fellowships, presentations, and notable outcomes.

VoltAge Doctoral Fellowship

Competitive research fellowship supporting PhD work on multi-agent AI systems for climate change mitigation through power grid optimization.

Concordia University • 2025

MITACS Accelerate Scholar

Research funding for Master's thesis on memory mechanisms for coherent LLM-based dialogue systems.

MITACS Canada • 2023–2024

Blacklight Summit 2025

Presented technical architecture of SpokenWeb Search Engine at the premier conference for digital library search systems.

Conference Presentation • 2025

IEEE CAI 2026 Publication

SWin paper on sliding window summarization for coherent LLM-driven dialogue systems accepted at IEEE Conference on Artificial Intelligence.

Granada, Spain • May 2026

EACL 2026 Publication

Accepted paper on causal bias mitigation in LLMs at European Chapter of the Association for Computational Linguistics.

Rabat, Morocco • March 2026

Connect

Open to research collaborations and consulting opportunities.

Let's Build Something Together

Interested in multi-agent systems, LLM reliability, or energy-aware AI? I'd love to hear from you.