Software engineer / Product lead

Chetan Singh

I take ideas from a blank page to commercial products. Backend systems, applied language AI, and the product judgment to know what is worth building.

  • NowCTO and Product Lead, Pinak AI Solutions
  • RootsNLP and backend engineering, Germany
Portrait of Chetan Singh in a navy blazer
CTO, Pinak AI Solutions
M.Sc. Data and Knowledge Engineering
Scroll
01

The short version

I have spent my career inside young companies, building the first working version of things that did not exist yet.

That path runs from IoT hardware, through language models and a recruitment automation product, to the logistics platform I now lead as CTO. Along the way I started a fintech venture of my own, which taught me to think about a product from the balance sheet backward as well as from the code forward.

What ties it together is a habit. I start from what the customer and the operation actually need, make the technical call that serves it, and keep the result reliable enough to run a business on.

0Years building software, since 2017
0Companies and ventures, most of them early stage
0Degrees, in India and Germany
0Competition prizes before graduating
02

Currently building

June 2026 to present · CTO and Product Lead

Pinak AI Solutions

As CTO and Product Lead, I am responsible for the engineering and product direction of both platforms. The objective is a shipping system that is reliable, scalable and ready for commercial use, with every stage of a shipment traceable from dispatch to delivery.

How a shipment moves through the platform

  1. 01DispatchShipment enters the system with the data it needs to travel.
  2. 02TrackingStatus is captured and surfaced as the shipment moves.
  3. 03Data flowInformation reaches every part of the platform that depends on it.
  4. 04DeliveryThe loop closes with a clean, end-to-end record.

What I own

  • Technical architecture and backend systems
  • Product direction, from concept to shipped feature
  • The trade-off between customer needs, operations and growth

Stack

PythonJavaScriptNode.jsMongoDB

The goal is software that is robust, scalable and ready for commercial use.

03

The journey

CTO and Product Lead

Pinak AI Solutions · Part-time, remote

Leading engineering and product for an international shipping platform, with commercial scalability as the standing requirement.

Founder, financial strategy

Stealth fintech venture · Berlin, remote

Started an early-stage fintech company and worked on its financial strategy and first steps in business development.

Data Scientist and Backend Developer

100 Worte Sprachanalyse / Psychological AI · Heilbronn, Germany

  • Built a recruitment campaign product from scratch, pairing automated outreach with psychological profiling of candidates.
  • Shipped microservices, REST APIs and external integrations on Node.js, MongoDB and Python, delivered through GitLab and Jenkins.
  • Created NLP pipelines for profiling, text augmentation and controlled generation using spaCy, BERT, GPT-2 and PyTorch.

Data Science Intern and Thesis Researcher

Psychological AI · Heilbronn, Germany

Trained LSTM language models for generation and classification, then used GPT-2 to steer generated text by keywords, style and tone for a master's thesis. Added Elasticsearch search and explainability checks along the way.

Junior Software Developer

Alteration Innotech · Pune, India

First professional role: client applications, backend services, and APIs that control hardware and IoT devices.

04

Selected work

Psychological AI · 2021 to 2025

Recruitment campaign engine

A product that runs outreach campaigns for headhunters, from the first message through to the close. It uses automated LinkedIn messaging and a psychological profile of each candidate to shape the conversation.

Node.jsMongoDBPythonNLP

Master's thesis · 2020 to 2021

Steering text generation with GPT-2

Research into guiding a language model so the output follows chosen keywords, style and tone. Built for real uses such as marketing copy, stories and product reviews, and evaluated with explainability methods.

GPT-2PyTorchExplainability

2020

Explaining black-box models

A pipeline on synthetic data to see how partial dependence and ICE plots behave when features interact.

Pythonscikit-learn
View repository

2019

EU legal text classification

Text analytics on a slice of the EURLEX corpus, tackling heavily imbalanced classes with Naive Bayes and neural networks.

TensorFlowNLTKImblearn
View repository

2018 to 2019

Spotting malicious activity

A machine learning pipeline that finds patterns in user activity on a system and flags the malicious ones.

RRandom ForestsNeural nets
View repository

Bachelor's project · 2016 to 2017

Home automation controlled by hand gestures

An IoT system you operate by moving your hand. Particle Photon and Raspberry Pi hardware, with Android and web apps on top.

Particle PhotonRaspberry PiPythonPHPAndroid
05

Craft and toolkit

Backend and systems

Node.jsMERNREST APIsMicroservicesAPI integrationsAWS

Languages

PythonJavaScriptJavaC++RPHP

Language AI

NLPBERTGPT-2LSTMFine-tuningspaCyTransformersText generation

Machine learning

PyTorchTensorFlowscikit-learnPandasNumPyExplainability

Data and search

MongoDBMySQLElasticsearch

Delivery

GitGitLabJenkinsLinuxJupyter

How I work

  • Technical thinking and managerial thinking in the same head
  • Clear technical documentation, so others can build on the work
  • Plain communication between engineers, customers and operators
  • Analytical habits: question the data, test the assumption

Recognition

  • First prizeCode-o-Mania coding competition, Dr. Jivraj Mehta Institute of Technology, 2015
  • Second prizeScientific paper presentation, Dr. Jivraj Mehta Institute of Technology, 2014
06

Foundations

Education

M.Sc. Data and Knowledge EngineeringOtto von Guericke University Magdeburg, Germany2018 to 2022
B.E. Computer Science and EngineeringGujarat Technological University, India2013 to 2017

Languages

  • EnglishProfessional
  • HindiAlso spoken
  • GujaratiAlso spoken
  • GermanBasic

Outside work

I keep building software for fun, usually to solve a small problem I ran into. When I am away from the keyboard I read, run, meditate, make music and vlog.

Let's talk

Have something that needs to exist?

I am open to senior software engineering and product roles, and to conversations about building something new.

singhchetan9924@gmail.com