AI Enablement Team Lead · Riedel Communications · Düsseldorf

Nitish
Aggarwal.: AI Enablement Team Lead, building AI products and the ground a company needs to use them.

✦Ask the AI concierge

✦

EXPERIENCE

10+

Years of experience

FOUNDER

Companies co-founded

DELIVERED

15+

Digital products delivered

EXPERIENCE

10+

Years of experience

FOUNDER

Companies co-founded

DELIVERED

15+

Digital products delivered

// the map

A 60-second walkthrough of this site

Six stops, start to finish. Tap any dot to jump straight to that part of the site.

01Ask AIStart here 02About meAbout 03PlaybookMethod 04WorkPortfolio 05NotesNotes 06Say hiReach out

Start here01Ask AIAsk my concierge anything
About02About meData scientist turned AI product lead
Method03PlaybookMy recipe for a good Tech PM
Portfolio04Work & ProjectsWhat I have built and shipped
Notes05Notes & WritingsField notes and essays
Reach out06Say hiBook a call or say hi

01 · About me

Who I am.

The model is never the hard part. Making an organisation trust it is.

Proven impact

saved through data-driven procurement: €500K+saved through data-driven procurement
pricelist cycle, 3+ months to under one: 3× fasterpricelist cycle, 3+ months to under one
of IT tickets deflected since go-live: ~15%of IT tickets deflected since go-live
revenue grown as a founder: $70K → $400Krevenue grown as a founder

In brief

From data scientist to AI Enablement Lead, my job has always been the same at heart: turn ambiguous business problems into AI that teams adopt and trust, and now into a way of working the whole company can share. The data thread runs through everything: analyst first, then a founder running a business on forecasts and churn models, then a Master's and work in pharma and energy.
At Riedel Communications I lead the AI Enablement Team. We build the company's AI products, including a multi-agent helpdesk in daily use by 1,500 people, and we run the framework that lets every department use AI safely and without waiting on us.
The AI work stands on foundations I built first: a BI layer across the ERP landscape, and an automated pricing cycle that replaced a three-month Excel marathon.
Shipping in a German enterprise means privacy, compliance, and consensus are part of the product, so I design them in from day one.
Before tech, I co-founded two startups, I know first-hand what it takes to build, lead, and own the bottom line.

Top skills

Technical

GenAI: LLMs, RAG & multi-agent
Python & SQL
Cloud AI: AWS Bedrock & Azure AI

Leadership

AI product strategy & roadmapping
Cross-functional leadership
AI governance: EU AI Act & GDPR

See all skills & tools

Education

M.Sc. Data Analytics and Decision Science · RWTH Aachen University
B.Tech. Electronics · ITM University

My journey

A non-linear path, from founder to AI product leader. Select any role for the full story.

Let's get in touch

02 · Playbook

The recipe for a Tech PM.

How I think and operate, with proof links on demand.

Recipe № 01 · Serves everyone

A recipe by Nitish · from the Düsseldorf kitchen

A Technical
Product Manager.

Prep · years on both sidesServes · business, engineers, usersLevel · senior heat

Ingredients

8 · non-negotiable

01
Business fluency
I look for where the real value lies, and back my decisions with research and feasibility tests.
02
Technical depth
I shape the architecture with the team, weighing constraints, value, and ROI.
03
Discovery discipline
I orchestrate discovery with the team, so every decision traces back to a real user story.
04
Saying no, with a reason
Not everything can be built. I choose what moves the product forward, and explain why the rest waits.
05
Two languages
I speak boardroom and engineering, without leaning too far into either.
06
Staying past go-live
Delivery isn't where my job ends. Adoption is the number I watch.
07
Trust as part of the product
Security, privacy, governance, explainability. A product without them works, but isn't trusted.
08
Judgment on trade-offs
Speed vs quality, scope vs deadline. Making those calls came with time, experience, and the people around me.

Method

7 steps · in order

01
I learn the business before touching the product.
Real ROI starts with understanding the business: interviewing stakeholders, and knowing the value a product should bring before it's built. Every decision after that inherits its quality from this step. The BI foundation below was built exactly on that groundwork.
See it in action: Product Analytics and BI Foundation
02
I earn technical credibility, I don't borrow it.
Technical, for me, means making architecture trade-offs, questioning estimates, and knowing what a design choice costs in money, risk and maintenance. That trust is what every roadmap runs on. The helpdesk's multi-agent architecture is where I put it to work.
See it in action: Multi-Agent IT Helpdesk
03
I drive strategic decisions with the team, not for it.
Ten stakeholders will always have ten wishes. Together with the team, I dig into the research and build a working prototype, and from there I make the call on what's worth building next, aligned with everyone behind it. The IT helpdesk started exactly that way.
See it in action: Multi-Agent IT Helpdesk
04
I sit between functions, not beside them.
Most of my value lives in translation: business intent into technical constraints, and technical reality back into business language. It's how I keep both sides moving in the same direction. Seven departments on one pricelist cycle came from exactly that.
See it in action: Pricelist Transformation
05
I build trust into the product itself.
Users only rely on a product they trust. So how a system handles data, what it may decide, and when it must escalate are product decisions I make early, not fixes I add later. The internal assistant was designed around this from day one.
See it in action: Riedel Internal Assistant
06
I measure adoption and act on it.
Shipping a product is not the goal, people using it is. I track adoption with the team, share the numbers openly, and change course when they tell us to. The helpdesk's deflection rate is the number we live by.
See it in action: Multi-Agent IT Helpdesk
07
I make it easier, and safer, for others to build.
Delivering one product is only part of the job. The bigger lever is helping other teams deliver their own: a clear way to request a tool or bring an idea, tools that are already checked, and support to run their own projects. The other half is the governance layer, making sure every AI tool in use and every project running in the company meets the EU AI Act and is secure. That is what the AI Enablement Team is for.
See it in action: AI Enablement Team

NextLearn more about me

03 · Top projects

Two that mattered most.

The two projects with the biggest impact, in numbers and in adoption.

AWS BEDROCK · MULTI-AGENT · RAG

Multi-Agent IT Helpdesk

A production multi-agent system on AWS Bedrock serving 1,500 employees. Deflects ~15% of the L1 queue and answers in seconds instead of hours.

~15% of L1 tickets deflected since go-live in June 2026.
Tickets that reach IT arrive pre-filled with conversation context.

Read the full project

PROCESS DIGITISATION · SAP

Pricelist Transformation

Rebuilt how a global company creates its pricelists: from a 3-month Excel marathon across seven silos to a transparent digital cycle that ships in under a month.

3+ month cycle → under a month. ~92% reduction in effort.
Seven departments working off one transparent cycle, not seven silos.

Read the full project

See all projects

04 · Skills and tools

Skills and tools.

The tools and skills I bring as a Tech PM.

Product and delivery

Product strategy
Roadmapping
Discovery
PRDs
Prioritisation
Agile / Scrum
Product vision
User research
User stories
Backlog management
Go-to-market
KPIs and success metrics
A/B testing
Zero to one delivery
Stakeholder management
Change management
Jira
Confluence

AI and GenAI

LLMs
RAG
Multi-agent systems
AWS Bedrock
Azure AI Foundry
Copilot Studio
Generative AI
Prompt engineering
NLP
AI product strategy
Model evaluation
Microsoft 365

Data, cloud and technical

SQL
Python
Power BI
SAP
AWS
Azure
Data analysis
Business intelligence
Data modelling
Data quality
APIs
System integration
Solution architecture

Governance and leadership

EU AI Act
GDPR
AI governance
Responsible AI
Data privacy
Data governance
Security
P&L ownership
Unit economics
Business cases
Process design
Operations management
Team leadership
Strategic thinking

05 · Notes and writings

Lessons and long reads.

The short version of what I have learned and what I write about.

Pinned · Top 5 learnings

from 10 years building

01
Where money goes matters as much as what gets built.
Budgets are strategy documents: what a company funds is what it actually believes. I learned that at 23, the expensive way, and never needed the lesson twice.
02
Consistency is a system, not a promise.
Four outlets, 250+ SKUs, one expectation: every visit tastes like the last. Quality that depends on any single person being present is not quality, it's luck.
03
You don't digitise files. You digitise agreements.
Three months of pricelist Excel wasn't a spreadsheet problem; it was seven departments without one process. The alignment was the product. The tool just enforced it.
04
Data, then process, then intelligence. In that order.
Companies that start with the AI ship demos. Trusted data and digitised processes are what turn a model into a product people rely on.
05
A team ships AI. A company needs an operating model.
Three products proved the value and showed the limit. The real work started when I stopped building for departments and started making it safe for them to build.