Shubham Shakya

Hi, I'm

SHUBHAM SHAKYA

AI PRODUCT ENGINEER

Building autonomous systems that think and act.

Professional Journey

Education
Experience

AI Software Engineer

Buffalo Early Learning Lab

  • React, jsPsych
  • Audio Processing
  • Audio transcription & speaker labeling

State University of New York at Buffalo

M.S. in Computer Science
  • Machine Learning
  • Deep Learning
  • Computer Vision
  • Reinforcement Learning

AI Engineer

NeurIOT Labs

  • RAG Systems
  • Agentic AI
  • Multi-Agent Orchestration
  • Information Retrieval

AI Engineer

ByteLearn

  • Computer Vision(OpenCV)
  • Pytorch Deep Learning
  • Image Segmentation(Mask R-CNN)

Software Engineer

Samsung Research Institute Delhi

  • Enterprise Display Solutions
  • Smart Signage & Interactive Boards

Research Intern

IIT Mandi

  • Social Network Analysis
  • Graph Embeddings(SDNE, Node2Vec)
  • TensorFlow

NIT Kurukshetra

Bachelor of Technology in Information Technology
  • Software Engineering Foundations
  • Data Structures and Algorithms
  • Operating Systems

Featured Projects

ReactExpressMulti-AgentPrompt EngineeringATS Optimization

ResumeForge: AI-Powered Resume Tailoring

Three-agent pipeline (Tailor → Validation → Formatting) that scores output against ATS criteria and loops back until a quality threshold is met, then exports a clean single-page resume as PDF or DOCX.

View Repo →
PythonClaude APIMCP ServerOllamaPrivacy-Preserving AI

Email Insights: Hybrid LLM Privacy Architecture

Two-system pipeline where a local Ollama model processes raw email data on-device and writes sanitized signals to SQLite; a frontier model (Claude) then queries only the sanitized store via MCP — zero raw data exposure.

View Repo →
GPT-4LangChainSnowflakeSQLPython

Agentic Data Analyst & Text-to-SQL

Autonomous analytics agent that converts natural language to SQL, querying Snowflake via an intelligent caching layer using GPT-4.

View Repo →
CLIPChromaDBPythonSemantic Search

Adaptive Video Search Agent

Scalable video retrieval system with human-in-the-loop adaptive feedback, utilizing CLIP and ChromaDB for semantic search.

View Repo →
BARTHuggingFaceNLPPython

Multi-News Research Agent

Autonomous multi-document summarization pipeline extracting key insights using BART.

View Repo →
CLIPGPT-4VPyTorchPython

VisualVerse: Creative Art Agent

Multimodal AI that perceives visual art and generates poetic, context-aware captions.

View Repo →
VAEPyTorchPython

Generative Emoji Designer

Deep generative model (VAE) capable of designing novel, diverse emoji assets from scratch.

View Repo →
NetworkXNLPGraph MLPython

Research Trend Forecaster

Temporal analysis engine tracking research trends using graph modeling and NLP.

View Repo →