Open to AI Engineering, Full-Stack & Data Platform roles

Nihildev
Nandakumar

AI Engineer building full-stack data products Published NLP researcher · Berlin

From a 50-day founding sprint that closed a €500K seed round to peer-reviewed NLP research at LREC-COLING — I build AI systems that reach production and stay there.

Years in AI &
Data Engineering

€500K

Seed funding
facilitated

100K+

AI evaluations
per day, built

LREC
'24

Published
NLP researcher

✉ Get in Touch ⬇ Download CV

🤖 LLM & Agentic AI 🐍 Python · FastAPI ⚛️ TypeScript · React 🧮 pgvector · Postgres ☁️ AWS · GCP · Supabase 📊 dbt · Dagster · BigQuery

About

Engineer by training,
builder by nature

I started writing software in Chennai, migrated legacy websites to the MERN stack, and visualised IPL cricket data in Tableau. That led to Munich — an M.Sc. at TUM, where four years of NLP research on differential privacy in language models ended up peer-reviewed and published at LREC-COLING 2024. Then Berlin: Infineon Technologies, where I ran NLP models and BI platforms for 100+ employees; Asgoodasnew, where I built ML pricing that moved margins; and finally PLACED, where I joined as Founding Engineer at an AI talent-matching startup — launched the founding product to early adopters in 50 days and helped close a €500K seed round.

I’m open to AI Engineer, Full-Stack Engineer, Data Platform Engineer, and Analytics Engineer roles — building systems that scale, ship, and show up in the numbers. Based in Berlin; open to Germany, EU, and remote. Let’s talk.

Quick Facts

📍

Location

Berlin, Germany

Open to remote / hybrid · EU

🎓

Education

M.Sc. Informatics, TUM

Technical University of Munich

📄

Research

LREC-COLING 2024

Differential Privacy in NLP

🏢

Most Recent Role

PLACED · Founding Engineer

Apr – Dec 2025 · HR-Tech startup

🌐

Languages

English · German · Tamil

Professional working proficiency

insights Discovery®

Value to the Team

Professional personality profile — how Nihildev shows up at work

Energy & Character

Brings boundless energy to any situation Knows nothing is impossible Life and soul of the group Provides stimulation and resourcefulness

Leadership Style

Leads by fighting alongside the troops Persuasive, inspirational leader Promotes ideas to, with and through others Initiates, justifies, actions and defends

Work Style

Challenges methods he feels are inadequate Communicates clearly and effectively

Selected Projects

Production-grade AI
built from scratch

Two flagship projects — each a full-stack AI system solving a real problem, built with production-engineering discipline.

AI Application · Full-Stack

Job Pipeline AI

Autonomous job search, ranking, and auto-apply

Upload your CV, set preferences — the system scrapes 15+ portals in parallel, ranks matches via pgvector hybrid retrieval, iteratively optimises your CV against top matches using GPT-4o in a LaTeX rewrite loop, and auto-applies via Playwright. All tracked in a real-time Next.js dashboard.

15+

Portals scraped

<1s

Match latency

GPT-4o

CV optimiser

Next.js 16 FastAPI pgvector Supabase Playwright Redis · BullMQ Docker OpenAI

View Presentation ↗ GitHub ↗

MCP Server · AI Data Engineering

DataMCP

Your AI Data Engineer, living in your editor

A production-ready MCP server that gives Claude Code, Cursor, and Codex the power to act as a full-stack data engineer. Describe analytics needs in plain English — DataMCP maps your Postgres schema, generates a complete dbt project, wires up Metabase, schedules Dagster pipelines, and provisions Kafka streaming. Backed by a persistent knowledge base that compounds with every session.

MCP tools

232

Tests passing

<15m

DB → dbt project

Python FastMCP dbt Dagster Apache Kafka Metabase SQLite KB Jinja2

View Presentation ↗ GitHub ↗

Impact in Numbers

The numbers that matter.
Earned, not claimed.

Days: founding MVP
to early adopters

95%

NLP model
accuracy achieved

€100K

Annual savings from
Tableau platform

80%

Manual effort cut
via AI automation

12%

Margin lift via
ML pricing engine

90%

Fewer deployment
errors via CI/CD

45%

Privacy protection
improvement (thesis)

30+

Developers reached at
AI Agents workshop

n8n-style

13 years of learning, building, and growing in tech.
One discipline: building what matters.

2013

2017

2019

2023

2025

2026

⌄ during M.Sc. · Munich

Education

2013 – 2017

B.Tech — Computer Science

Software Engineer

2017 – 2019

SportsMechanics India

Education

2019 – 2023

M.Sc. Informatics · TUM

Web & Data Specialist

2023 – 2025

Asgoodasnew Electronics

Founding Engineer

2025

PLACED · AI/HR-Tech

AI Engineer

GTM

2026

HOLIDU

Data Analyst

2019 – 2023

Infineon Technologies

Master Thesis

LREC-COLING 2024

NLP Privacy Thesis

n8n workflow

Career Journey

Seven years. Three countries.
One consistent thread.

From scraping cricket data in Chennai to architecting an AI platform that closed a seed round in Berlin — every role moved further along the same trajectory.

Apr 2025 – Dec 2025 📍 Berlin, Germany

Founding Engineer

PLACED Startup · AI / HR-Tech

50-day founding sprint: Shipped the startup MVP and onboarded early adopters with AI-first development (Copilot + Cursor) — the fastest product launch of my career.
Agentic matching pipeline: Multi-context LLM orchestration reaching 100K+ evaluations per day at sub-second latency on AWS.
€500K seed round: Technical demonstrations were the centrepiece of the fundraise. Owned the full stack: Python/FastAPI backend + TypeScript/React frontend + CI/CD.

PythonFastAPITypeScript ReactAWSLLM Orchestration GitHub ActionsSupabase

Nov 2023 – Jan 2025 📍 Berlin, Germany

Web & Data Analytics Specialist

Asgoodasnew Electronics Scale-up · E-Commerce

ML pricing engine: Competitor price monitoring model that improved profit margins by 12% and cut manual tracking by 80%.
ELT automation: Rebuilt workflows with GCP Dataform — 50% faster processing, 20% cost reduction.
Server-side tracking: GTM server-side implementation increased event accuracy by 40%; A/B experiments drove measurable conversion uplift.

BigQueryPythonGCP Dataform GTMML Pricingdbt

Dec 2019 – Sep 2023 📍 Munich, Germany

Data Analyst

Infineon Technologies Enterprise · DAX 40

NLP sentiment model: 95% accuracy — eliminated 80% of manual feedback classification across the organisation.
Tableau BI platform: 30+ executive dashboards serving 100+ employees, generating €100K in annual savings. Won the #oneSAP Team Award.
Predictive sales: Forecasting models that grew profitable customer segments by 15%.

PythonNLPPyTorch TableauSQL

Sep 2017 – Mar 2019 📍 Chennai, India

Software Engineer

SportsMechanics India Sports Tech

MERN migration: Moved 10+ client websites to React/Node stack, cutting maintenance costs by 30%.
IPL dashboards: Real-time cricket analytics reducing data processing time by 50% — where the data obsession started.

ReactNode.jsMongoDB PythonTableau

Research & Thought Leadership

The rare engineer who
publishes what they build

Peer-reviewed at a flagship NLP conference. Workshop speaker to 30+ developers across Europe.

📄 Published Paper

LREC-COLING 2024 · Torino, Italy

Comparative Analysis of Local Differential Privacy in Natural Language Processing

A rigorous comparative study evaluating local differential privacy mechanisms applied to language models. Demonstrated a 45% improvement in privacy protection while maintaining model utility — a core tension in privacy-preserving AI. Published from my M.Sc. thesis at TUM, contributing novel empirical findings to the field.

🏛️ Technical University of Munich 📅 May 2024 🔬 NLP · Privacy · ML

Read on ACL Anthology ↗

🎤 Workshop Speaker

International Showtimes · Europe · April 2025

AI Agents & Automation: Building the Next Wave of Enterprise Intelligence

Designed and delivered a full-day workshop on AI agents and automation to 30+ developers. Topics: LLM multi-step orchestration, prompt engineering, n8n automation. Sparked adoption of no/low-code AI tools in production workflows — part of International Showtimes' Q2 initiative to become the most productive team on the planet.

👥 30+ Developers 📅 April 2025 ⚙️ LLMs · Agents · Automation

Education

Academic foundations
that still pay dividends

🇩🇪

M.Sc. Informatics

Technical University of Munich (TUM)

Jan 2019 – Jan 2023

Specialisation in NLP, ML, AI, and Software Engineering. Thesis on Local Differential Privacy in NLP — published at LREC-COLING 2024, with 45% privacy improvement while maintaining model utility.

Grade: 2.0 · German Scale

🇮🇳

B.Tech Computer Science

SRM University, Chennai

Jan 2013 – Jan 2017

Software development, data structures, and algorithms. First Class with Distinction — top 5% of the graduating class.

GPA: 9.3 / 10 · Top 5%

Tech Arsenal

The full stack,
layer by layer

AI & Machine Learning

LLM OrchestrationPrompt Engineering Agentic AI SystemsRAG Pipelines OpenAI GPT-4oEmbeddings & pgvector PyTorchNLP (spaCy · HuggingFace) Differential PrivacyMCP Servers

Backend & APIs

Python 3.11+FastAPI Node.jsTypeScript Redis · BullMQPostgreSQL SupabaseSQLAlchemy

Frontend

React 19Next.js 16 App Router TailwindCSSZustandTanStack Query

Data Engineering & Analytics

dbtDagsterApache Kafka dltBigQueryGCP Dataform MetabaseTableau

Cloud & DevOps

AWS (EC2 · S3 · Lambda)GCP DockerGitHub Actions (CI/CD) Infrastructure as CodeRailway

AI-Assisted Development

Claude CodeGitHub Copilot Cursorn8n Automation

Testimonials

What colleagues
and founders say

We kicked things off with a great workshop led by Nihildev Nandakumar on AI agents and automation. One key takeaway: the space is evolving fast, especially with no/low-code tools like n8n, which we started using right away… it sparked valuable discussions and laid the groundwork for long-term impact.

You are an incredible developer and your family can be extremely proud of you! Let's transform this industry together and bring candidate placement into the era of AI!

The oneSAP Training Dashboard built with Nihildev tracks the complete training process — invitations, registrations, KPI analysis for go-live decisions, trainer feedback. Automated daily updates. Supports multiple simultaneous rollouts across Infineon's global digitalization strategy.

NihildevNandakumar

Engineer by training,builder by nature

Production-grade AIbuilt from scratch

The numbers that matter.Earned, not claimed.

13 years of learning, building, and growing in tech.One discipline: building what matters.

Seven years. Three countries.One consistent thread.

The rare engineer whopublishes what they build

Academic foundationsthat still pay dividends

The full stack,layer by layer

What colleaguesand founders say

The right hireis one message away

Nihildev
Nandakumar

Engineer by training,
builder by nature

Production-grade AI
built from scratch

The numbers that matter.
Earned, not claimed.

13 years of learning, building, and growing in tech.
One discipline: building what matters.

Seven years. Three countries.
One consistent thread.

The rare engineer who
publishes what they build

Academic foundations
that still pay dividends

The full stack,
layer by layer

What colleagues
and founders say

The right hire
is one message away