Projects

Dialectic is an adversarial reasoning application built to improve decision quality on complex questions by forcing two AI agents to debate opposing sides instead of generating a single one-shot answer. GitHub: github.com/singhaditya8499/Dialectic What We Built We built a complete end-to-end debate platform with: Two-agent structured debate loop (for vs against, or A vs B) Provider-agnostic model layer supporting OpenAI, Anthropic, and Ollama Live streamed debate transcript over SSE for real-time UX Evidence-anchored turn schema (facts, figure/date, source, reliability) Structured summary pipeline that preserves major claims and evidence ledger Local debate archive with reload support from the UI Browser-compatible mode for GitHub Pages deployments Product and Technical Design Core architecture src/server.js: HTTP API, SSE stream endpoint, static asset serving, persistence APIs src/debateEngine.js: debate orchestration, turn normalization/validation, summary generation src/providers.js: common completion abstraction across providers public/*: setup form, transcript renderer, summary view, saved debate browser Debate lifecycle User submits question, side models, and round budget. Engine alternates turns between both sides. Each turn is validated for stance, rebuttal quality, novelty, and evidence quality. Streaming events update UI in real time (thinking, turn, summary, complete). Final structured summary and debate artifact can be saved locally. Quality and Alignment Guardrails Dialectic focuses on controlled generation rather than raw free-form output: ...

This book chapter proposes an AnoGAN-inspired approach tailored for tabular datasets, targeting improved anomaly detection performance beyond common baselines. Publication link: Read on Springer

This article explores BK Trees and how they enable efficient approximate string matching using edit distance. Publication link: Read on Medium

CardWise AI is a personal-finance assistant focused on helping users evaluate credit-card options, align card selection with financial goals, and understand spending behavior through AI-generated summaries. Focus Areas Personalized card recommendations based on user goals and constraints. Goal-aware planning support for safer and more structured decision-making. Natural-language insights to explain spending patterns in simple terms.

Customer Reise is a support-focused web application built to reconstruct and replay customer journeys from logs so teams can diagnose user-reported issues faster and with more confidence. Highlights Aggregates cross-service logs into a coherent user journey timeline. Enables issue replay workflows for support and engineering teams. Improves root-cause analysis speed for production incidents.

Overview DocCompass is a local-first document discovery system built to find personal files by meaning, not only by filename. It is designed for natural requests like: “my marksheets” “visa related documents” “intro to psychology notes” The project searches local directories, extracts text from multiple file formats, builds a reusable index, creates a structured query plan, and ranks results by relevance with transparent scoring. Repository: github.com/singhaditya8499/DocCompass Problem Statement Traditional local file search is often fragile because it depends heavily on exact filenames and folder memory. In real usage, users usually remember intent and partial context, not exact paths. DocCompass addresses this by combining: ...

Face Recognition System is a practical computer-vision pipeline that covers full lifecycle stages: data capture, model training, and live identity recognition. Highlights Captures and curates face samples for training. Trains a recognition model from collected data. Runs real-time identification on live video streams.

This conference publication presents a filter-based blockchain architecture intended to improve scalability for domestic IoT environments. Publication link: Read on IEEE Xplore

Graph Plotter is a C++ graphics project focused on rendering mathematical functions at pixel level to build intuition for plotting logic and coordinate transformations. Highlights Plots trigonometric and polynomial functions. Handles axis mapping and coordinate conversion. Built with classic graphics.h style rendering primitives.

Hand Game is a real-time interaction project where hand gestures become game controls through computer-vision based finger and contour detection. Highlights Uses webcam input for gesture capture and detection. Converts hand movements into gameplay actions. Demonstrates low-latency vision-to-input integration.

LeetCode Rank Finder is a lightweight automation utility that solves the practical challenge of locating a specific contestant quickly in large contest ranking tables. Highlights Fetches and parses contest ranking data. Searches standings by username and returns rank details. Reduces manual effort during contest analysis.

Markov Chain Basic is an exploratory NLP project that uses Markov transitions to generate text and study long-run behavior in probabilistic sequence models. Highlights Builds transition maps from input corpora. Generates synthetic text using stochastic state transitions. Serves as an intuition-first bridge into probabilistic NLP.

This publication explains Markov chains with an intuition-first approach and practical examples for state-transition thinking. Publication link: Read on Medium

This paper applies game-theoretic modeling to basketball shot-distribution strategy and examines team-level performance optimization. Publication link: Read on arXiv

An accessibility-focused ML prototype that translates blink patterns into actionable commands, designed to support learning interaction for users with physical limitations. Repository/demo assets can be embedded here as the project evolves.

A mental wellness web app mapping anxiety indicators from HAM-A scoring to personalized music recommendations using emotion-aware metadata and fast feedback loops. Repository/demo assets can be embedded here as the project evolves.

An AI-driven recommendation prototype combining LlamaIndex pipelines, multi-model analysis, and sentiment extraction from financial news streams. Repository/demo assets can be embedded here as the project evolves.