✳AI, technology & thoughtful experimentation

Building with AI. Understanding risk.

I’m Nitish — passionate about technology, building with AI, and understanding the risks around it. I assess technology risks and identify practical mitigation steps to help make systems more secure and reliable.

Noida, India/ Always learning
IN MY OWN WORDS

Meet the person
behind the projects.

A space for my story, interests and what I’m building next.

My video introduction

Coming soon — a personal hello.

Add introduction video ↗
THE INTERSECTION OF CODE & CURIOSITY01 / ∞
A LITTLE EXPERIMENT IN PERSPECTIVE
MY TOOLKITAlways evolving ↗
Governance, Risk & ComplianceSOC 1 & SOC 2 AuditsMCA AuditsIT General Controls (ITGC)AI Governance & Responsible AINIST AI RMFMachine LearningGenerative AIData AnalyticsAI Model ValidationAI Risk AssessmentExplainable AI (XAI)Deep LearningComputer VisionPythonTensorFlowOpenAI APIsLarge Language Models (LLMs)LSTMXGBoostFeature EngineeringAI Lifecycle ManagementAudit Automation
ENGINEERING × PERSPECTIVE

Machine learning ✳ Responsible AI ✳ Technology governance

01 / A LITTLE CONTEXT

A builder’s curiosity.
A broader perspective.

My journey spans machine learning development, data science and technology governance. I enjoy connecting the technical details of how a system works with the people, processes and questions around it. Here, I document personal experiments and selected experiences that have shaped the way I think.

02 / PERSONAL MILESTONES

Progress I’m proud of.

Education, experience and recognition
that have shaped my journey.

2024

MBA in Data Science & Data Analytics

Completed at Symbiosis Centre for Information Technology, SIU, Pune, building on an engineering foundation.

Career milestone

From AI development to technology governance

Experience spanning machine learning, computer vision, model development and technology control reviews.

Recognition

Collaboration Extraordinaire

Recognized for outstanding collaboration in supporting and delivering projects.

03 / PROJECTS

Ideas, brought
into focus.

Explore my projects and the ideas,
methods and learning behind each one.

Personal project · interactive walkthrough

TrustSignal

A guided gap-assessment project

A personal project that brings a guided gap assessment, control mapping and an evidence library into one workspace. Explore the full workflow and learn how the results are put together.

PythonFlaskGap assessment
Behind the project +

Start with a profile, select Trust Services categories, answer the control questions and review the illustrative results. The original TrustSignal workflow includes reports, exports, ControlMap and saved reviews. This is an educational self-assessment, not a commercial risk assessment, consulting service or audit opinion.

Start a gap assessment ↗

Personal project · runnable local demo

KYC system for identity verification

Document OCR and face comparison on your computer

Run the full KYC workflow with PAN or Aadhaar images, local OCR, exact field checks and selfie comparison. Start with a fictional sample or explore the detailed project walkthrough.

PythonFlaskOCROpenCV
Behind the project +

The current repository runs RapidOCR with ONNX Runtime, YuNet face detection and SFace embeddings on the CPU. Select PAN, Aadhaar or both, compare document text, then capture or upload a selfie. Sample controls use the same real model pipeline. Results show consistency under the configured rules, not document authenticity or liveness. Includes complete workflow, code structure and setup instructions.

Try the demo ↗View source on GitHub ↗

Personal project · runnable local demo

DocuFind AI

Find answers across approved company documents

A runnable Python RAG chatbot demo that retrieves relevant document passages, answers questions from that evidence and shows source references.

PythonRAGLLMDocument search
Behind the project +

The updated project includes a browser chat, a local document library, Markdown and plain-text uploads, passage retrieval, source citations and an optional OpenAI answer path. It runs locally without third-party packages in local-answer mode. The demo uses lexical retrieval and is intended for learning; it is not a production access-control or document-security system.

Open project ↗View source on GitHub ↗

Internship experience · detailed project walkthrough

ITOps Root cause analysis with ML

From incident tickets to seasonal anomaly detection

Explore my 2023 internship project through 13 detailed steps and 21 report figures: ticket similarity, resolution scoring, LSTM forecasts and anomaly detection.

NLPBiLSTMXGBoostTime series
Behind the project +

A report-based educational case study covering data preparation, cosine similarity labels, bidirectional LSTM, sentiment experiments, XGBoost resolution ranking, seasonal forecasting and dynamic anomaly thresholds. Includes original figures, recorded results, evaluation limitations, lessons and the internship timeline.

Read the full project ↗

Experience entries describe past work. The ITOps case study uses the supplied internship report and its technical figures; underlying datasets, trained models and employer applications are not included.

04 / THE JOURNEY SO FAR

Different chapters.
Connected interests.

From building models to thinking about
the systems around them.

May 2024 – Present

Consultant · Technology Risk

Ernst & Young Associates LLP

At EY, I assess technology risks and review controls across enterprise applications, cloud platforms and AI-enabled workflows.

April 2023 – June 2023

AI Intern

UST Global

Explored machine learning for IT operations through data analysis, feature engineering and predictive modeling.

October 2018 – March 2022

AI/ML Engineer

WebOccult Technologies

Worked across machine learning, deep learning, computer vision and the AI lifecycle, from data preparation to model validation and deployment.

05 / FOUNDATIONS

Learning that
keeps evolving.

2024

MBA · Data Science & Data Analytics

Symbiosis Centre for Information Technology, SIU, Pune

2019

BE · Electronics & Communication

Gujarat Technological University, Ahmedabad

Languages English · Hindi · Gujarati

06 / TRY MY PERSONAL PROJECT

How to run a gap assessment
with TrustSignal.

Start a gap assessment ↗

A guided walkthrough from your first answer to an illustrative report. Open TrustSignal to use the complete assessment, ControlMap and evidence library.

  1. 01

    Set the context

    Enter a project alias and choose the organization type, report type and planning horizon. Use sample or sanitized information.

  2. 02

    Choose the scope

    Security is included. Add the other Trust Services categories relevant to the scenario you want to explore.

  3. 03

    Answer the questions

    Work through the control questions. Choose implemented, partial, not implemented or not sure, and read the example evidence.

  4. 04

    Explore the results

    Review the illustrative score, gaps and suggested actions. Download the report or CSV, or explore ControlMap and evidence library.

07 / KEEP THE CONVERSATION GOING

Good ideas start
with a hello.

For conversations about learning,
technology and shared interests.

Say hello ↗LinkedIn ↗

jhanitishkumar12@gmail.com