IIT Delhi Graduate • 20+ Projects, AI/ML Models & Audio Plugins

I Build AI Systems & Software Products That Drive Real Results

For AI startups and founders who need to automate workflows, launch smarter products, or build revenue-relevant software — I design, build, and ship end-to-end AI and software systems.

20+
Projects Delivered
5/5
Client Rating
IIT Delhi
Background

Technologies

Tech Stack & Tools

JavaScript
TypeScript
Python
C++
TensorFlow
PyTorch
OpenCV
Scikit-learn
NumPy
Pandas
FastAPI
Flask
Node.js
React
Next.js
Tailwind
Framer Motion
PostgreSQL
Redis
Docker
Git
Selenium
J
Java
K
Keras
C
Celery
n
n8n
S
Supabase
P
PyQt6
J
JUCE
L
Librosa
K
Keras
J
Java
Selenium
Git
Docker
Redis
PostgreSQL
Framer Motion
Tailwind
Next.js
React
Node.js
Flask
FastAPI
Pandas
NumPy
Scikit-learn
OpenCV
What I Build

Focused on Four Core Areas

Each service area is built around real business problems — not theoretical capabilities.

AI/ML Recommendation Systems

Personalization engines, collaborative filtering, and content-based AI that improve user retention, increase engagement, and surface the right content at the right time.

Recommender SystemsPersonalizationRanking

FinTech ML

Fraud detection, credit scoring, and anomaly detection models that protect revenue, reduce risk, and automate decisions that would otherwise require manual review.

Fraud DetectionCredit ScoringAnomaly Detection

Audio Intelligence Pipelines

Commercial-grade VST/AU plugins, real-time audio analysis, and AI-powered audio processing systems for music tech companies and production environments.

VST/AU PluginsAudio MLDSP

Full-Stack AI-Backed Products

End-to-end software products that combine AI/ML backends with polished frontends — built to production standards and designed to ship fast.

Full-StackAI ProductsSaaS
5/5 Average Rating
20+ Projects Delivered
IIT Delhi Background
Signed Project Agreement
Founder Product

NomoCrit: AI Decision Accountability Infrastructure for Lending

NomoCrit is a founder-built AI governance layer for Indian NBFC and digital lending systems: making credit decisions explainable, auditable, and RBI-aligned in real time, not retrospectively.

Built end-to-end: explainability engine, adaptive threshold recalibration, fairness monitoring, and audit-trail infrastructure. Demonstrates the ability to architect, build, and ship regulated fintech infrastructure from zero — not just a model, a compliance substrate.

Visit NomoCrit
Explainability Engine
Every credit decision is decomposed into a per-feature SHAP contribution map, surfaced as machine-readable JSON and a human-readable grounds-of-rejection notice.
Adaptive Threshold System
Approval thresholds recalibrate automatically when population-level drift is detected across a rolling credit window — without manual model retraining.
Fairness Monitoring
Demographic-parity and equalised-odds metrics computed on every decision batch, with automated alerts when protected-class disparity exceeds tolerance bands.
Audit-Trail Infrastructure
Decisions, explanations, threshold states, and fairness metrics written to an append-only ledger at evaluation time — export-ready for RBI inspection.

Selected Work & Delivered Results

Commercial products, AI systems, and software tools built for real business outcomes.

Live

Music Intelligence Platform — Alenka Media

AI system that automates content classification and metadata enrichment for a music tech client. CNN-LSTM hybrid achieves 89% accuracy, enabling the platform to tag and categorize music at scale without manual review.

PythonTensorFlowFastAPIWhisperCNN-LSTM
Live

NomoCrit — AI Governance Infrastructure for Lending

Founder-built AI governance layer for Indian NBFC lending — real-time explainability, adaptive threshold drift detection, and audit-trail infrastructure for RBI-aligned credit decisions.

PythonSHAPFastAPINext.jsPostgreSQL
Production Ready

Scale Lock — Commercial VST/AU Plugin

Production-grade MIDI intelligence plugin used by music producers in professional DAWs. Real-time scale mapping with timeline automation eliminates a key workflow friction point — shipped to a paying client and in active use.

C++JUCEVST3AUReal-time DSP
Production Ready

FootstepDetector — AI Audio Plugin

AI-powered VST/AU plugin for game audio and film production that detects and isolates footstep sounds in real time using neural network inference — automating a task that would otherwise require manual audio editing.

C++JUCELibTorchVST3AU

What I Build & How I Build It

Applied ML, audio engineering, and full-stack product work — focused on software that creates business value.

Audio Intelligence

  • VST/AU plugins for commercial music production
  • Real-time MIDI processing & musical intelligence
  • Audio ML: classification, analysis, feature extraction
  • DSP for effects, synthesis, and sound design
  • AI-powered audio analysis pipelines

AI & Machine Learning

  • Recommendation and personalization engines
  • Fraud detection and anomaly detection systems
  • NLP for content analysis and data extraction
  • Classification models for business decisions
  • Time-series and predictive analytics

Product Engineering

  • AI-backed web products: React, Next.js, TypeScript
  • Python APIs (FastAPI) and Node.js services
  • Database design: PostgreSQL, MongoDB, Supabase
  • Production deployment: Docker, Railway, Vercel
  • End-to-end: design → build → ship

What Clients Say

Real results for founders, startups, and product teams across music tech, AI, and software.

"Advait is one of the best JUCE programmers out there. He is conscientious, pays attention to detail and goes beyond the call of duty – and beyond the call of the specs – if needed to deliver a professionally well-programmed result. Highly Recommended!"

Akiva Shapiro

Founder, ChordBoard Music (US)

VST3/AU Plugin Development

"Very great work will surely be working with him again and very professional and attention to detail"

Miguel Magri

CEO, Tunem (Malta)

Deep Learning Implementation

"Understood our requirements, suggested improvements, and followed through on them. Would certainly work with him again."

Anish TrivediREPEAT

CEO, Alenka Media (India)

Deep Learning & ML Systems

"Really went above and beyond with extra edits to the project even after the work was done. would definitely recommend and will be working with him again."

Anonymous ClientREPEAT

Audio Production Professional (UK)

Audio Plugin Development

"Advait was a pleasure to work with. professional, skilled, and timely. He took feedback well and delivered excellent work to high standards."

Anonymous Client

Music Software Company (US)

Audio Plugin Development & Integration

5/5

Average Rating

20+

Projects Delivered

Let's Build Together

Have a Project in Mind?

AI startups and founders — I help you build the systems that move the needle.

Typical response within 24 hours.