I spent 12 years in data. Now I build AI systems that put it to work.

- 12 years
- in data and analytics, from startups to enterprise
- 15+ companies
- where I've led workshops and projects
- 500+
- data people in the DnA community I run
What data work taught me about AI
The habits from my data years are the ones that make AI systems useful.
Most AI problems are really data problems.
The model is rarely what fails. Missing context, unclear definitions and messy access are. So that is where I start.
Start with the decision, not the tool.
If you can’t name who will act differently because of it, it isn’t ready to build. I pick the smallest set of tools that does the job, often ones you already pay for.
Agree on definitions before answers.
AI makes inconsistent definitions louder, not quieter. Writing down the core business definitions first saves months of arguing with confident but conflicting answers.
What each stage of company taught me
The same AI system needs a different path in a ten-person startup and in an enterprise.
Startups
You get one shot with little time and less data. I learned to find the one number that changes a decision.
Scale-ups
Growth breaks things quietly. Definitions drift, tools pile up and nobody trusts the dashboard.
Enterprise
Good work dies without buy-in. Governance, security and many stakeholders decide what actually ships.
How I work with clients and their teams
More than anything, I'm a people person. A system only succeeds when the people who use it trust it and help shape it, so I spend as much time with your team as with your data.
I only recommend what I have built and run myself, and I measure success by whether the team can run and improve the work without me in the room. Over the years that has taken three different forms.
- Workshops
- Hands-on sessions that taught a team new ways of working with AI. They are part of my experience, not a current offer.
- Advisory
- Leadership or a team does the work, and I advise on the decisions and the implementation.
- Hands-on projects
- I lead the build until the solution works in the client’s environment, with their team involved throughout.
Companies whose teams I've worked with through workshops, advisory work or hands-on projects. A logo is not an endorsement of a specific result.
What clients say
“Beyond introducing agentic workflows, he followed up closely with both me and the team to ensure we implemented things the right way, both tactically and strategically.”
Read the full quote from Ran Almog
“Working with Nimrod started with an AI workshop for our data teams and quickly evolved into a real shift in how analysts work day to day. Beyond introducing agentic workflows, he followed up closely with both me and the team to ensure we implemented things the right way, both tactically and strategically. The frameworks we built together became part of the team's new normal and created a lasting impact on our analysis processes.”
Ran AlmogHead of Data, Elementor“He consistently helps us focus on practical, high-impact use cases and avoid common pitfalls, enabling us to move faster and make better-informed decisions about our AI strategy.”
Read the full quote from Assaf Levnison
“Working with Nimrod has helped us ensure we're approaching AI for marketing analytics using proven best practices and industry-leading approaches. His expertise in agentic workflows, context management, and AI system design provides valuable guidance as we evaluate and build our framework. He consistently helps us focus on practical, high-impact use cases and avoid common pitfalls, enabling us to move faster and make better-informed decisions about our AI strategy.”
Assaf LevnisonDirector of Marketing Analytics, Gong“The system our DA team built with Nimrod is one of the more sophisticated agent architectures we've deployed -- multiple agents working in concert across the entire analytical pipeline, triggered automatically when a production issue surfaces.”
Read the full quote from Yotam Galon
“The system our DA team built with Nimrod is one of the more sophisticated agent architectures we've deployed -- multiple agents working in concert across the entire analytical pipeline, triggered automatically when a production issue surfaces. It dramatically accelerated the workflow and delivered measurable impact on resolution speed and revenue risk. The kind of work that raises the bar.”
Yotam GalonVP Operations, Bigabid“Beyond the strategy, he mentored our analysts hands-on and left us with frameworks the team still uses.”
Read the full quote from Itay Aizik
“Nimrod works with our data group for several months and gave us a clear direction on how to transform the way we work with data using AI. Beyond the strategy, he mentored our analysts hands-on and left us with frameworks the team still uses. He consistently went above and beyond what we asked for - always available, always with a smile, and with a level of professionalism that's rare to find.”
Itay AizikData Platform Team Lead, ArtlistThe DnA community
I run DnA, a community of more than 500 data people. It keeps me close to the people doing data work every day.
Visit the DnA community (opens in a new tab)Open source and public resources
Some of what I build I publish for anyone to use, alongside AI Analytics Hub, my free resources on AI and data.
- Agent Think MapYour agent took 40 steps. Find the one that matters.
- Data Context Layer StudioGive your AI agents a governed context layer.
- Data Analytics SkillsAn open-source Claude and Codex skills library.
- AI Analytics Hub — by Nimrod FisherArticles, guides, directories and open-source projects. Free to read.
- Weekly updatesA roundup of AI releases across data platforms, published most weeks. Free, no sign-in.
Looking to solve your next big business challenge with AI and Data? Let's talk it through.
Tell me how we can make your business grow faster. I'll tell you honestly what I would build, and what I wouldn't.
Gong
Bigabid