Hi, I'm Samiksha
Building intelligent AI agents that remember and reason.
SS

About

AI/ML Engineer passionate about building intelligent systems. Currently working on multi-agent workflows and building ContextMemory, a memory system for AI applications. I specialise in Machine Learning, Deep Learning, NLP, and backend development with Python and FastAPI.

Skills

Machine Learning
Deep Learning
PyTorch
Agentic AI
Multi-agent Workflows
Langchain
LangGraph
Pydantic AI
LangSmith
Giskard
RAGs
Memory Systems
FastAPI
SQLAlchemy
Pydantic
Streamlit
My Projects

Check out my latest work

I've worked on a variety of AI/ML projects, from memory systems to workflow automation platforms. Here are a few of my favorites.

ContextMemory

A memory system for AI apps that remembers context across conversations. Extracts and stores facts, uses semantic search for retrieval, handles deduplication intelligently. Available as an open-source pip package.

Python
OpenAI
PostgreSQL
Semantic Search
Vector Database
RAG
Graph DB
PyPI

AIxAI

MCP-powered system that generates Pydantic AI agents using multi-agent LangGraph workflows. Crawls and chunks docs, stores in Supabase with OpenAI embeddings. Provides semantic search and RAG-based answers.

Agentic AI
MultiAgent Workflow
Pydantic AI
LangGraph
Supabase
Agentic RAG
MCP
Python

HomeSage

Smart companion for real estate price prediction and analytics. Leverages ML for accurate predictions and personalized property recommendations.

Python
Machine Learning
Data Analytics
Data Visualization
Data Engineering
Feature Engineering
EDA
Prediction Models

n8n

Visual workflow automation platform built with React and FastAPI. Design, connect, and automate workflows in real time with a drag-and-drop interface.

React
FastAPI
Pydantic
SQLAlchemy
Python
Workflow Automation
Real-time

Transformers Research

Deep dive into 'Attention Is All You Need' paper. Implementing transformer architecture from scratch, studying self-attention mechanisms and positional encodings.

PyTorch
Transformers
Self-Attention
NLP
Deep Learning
Research

Tiny Recursive Model

Researching efficient recursive model architectures and parameter-efficient approaches for LLMs. Exploring recursive computation and memory optimization in neural networks.

PyTorch
LLMs
Model Optimization
Recursive Networks
Deep Learning
Research
Hackathons

I like building things

Solving real-world problems with innovative tech, from IoT systems to blockchain applications.

  • S

    Smart India Hackathon (SIH)

    India

    Represented 6-member team in designing IoT-based Sewage Problem Alert system, securing top finalist position among 30,000+ participating teams. Developed a comprehensive solution for real-time sewage monitoring and alerting.
  • H

    Hack JKLU

    India

    Conceptualised and prototyped Blockchain-based eVault system with 256-bit encryption, earning 3rd place recognition for innovation. Built a secure digital vault solution leveraging blockchain technology for enhanced data protection.
Contact

Get in Touch

Want to chat? Just shoot me a dm with a direct question on twitter and I'll respond whenever I can. I will ignore all soliciting.