Shaping EcosystemsScaling Platforms
An agent is only as good as the context it can reach, and in industry that context sits in connected products and their platforms. I have helped shape and sell those platforms since 2001, in the Bosch group from 2015 to 2026. Today I give manufacturers and investors vendor-neutral advice, never bidding for the build, and help platform companies and startups sell to industry.
Vendor-neutral advice on AI agents, IoT and go-to-market for manufacturers, platform companies, startups and investors. Helping shape and sell platforms since 2001, at Bosch from 2015 to 2026.
As of October 20263–9 months or fixed feeGevelsberg (DE) · Athens · EUGerman · Englishkai@kaihackbarth.com

Each is a mandate of its own and follows the same order: first the decisions that get expensive later, then the build. The simplest start is a four-week fixed-fee audit, with me directly.
For Manufacturers and industrial companies, from the upper mid-market to large enterprises
Operating model first: who owns the agent, what counts, where a human decides. Then build vs buy, choosing the partner and governing them through to sign-off.
For Makers of connected products, from machinery to home and building technology
Foundation first: who owns the fleet, how a fix reaches every device, and what happens when the cloud, the module or the network reaches end of life before the product does. Agents on the shop floor later run on the same foundation.
For IoT and AI platform companies, industrial-tech scale-ups and startups
Market and use case first, then price and a sales path others can repeat, from selling platforms across industries myself.
For investors Every audit also as commercial or tech due diligence, then interim mandates in the portfolio.
To this day, Kai remains my absolute go-to advisor for understanding how IoT works at scale.
“I had the distinct pleasure of collaborating with Kai on a global open-source IoT project. He brings the kind of leadership and deep technical expertise that can only be forged through years of building large-scale IoT solutions. He was truly the face of Bosch’s IoT initiatives. To this day, Kai remains my absolute go-to advisor for understanding how IoT works at scale. A true visionary in the space — I cannot recommend him highly enough.”LinkedIn · May 2026
Kai has the rare ability to maintain a big-picture perspective while mastering technical details …
“Kai has the rare ability to maintain a big-picture perspective while mastering technical details, making discussions both strategic and highly practical. He doesn’t just follow trends — he actively contributes to them, bridging the gap between industry players, technical experts, and business needs. I wholeheartedly recommend him for any project involving IoT solutions, Edge computing, or environments that require both technical insights and strategic vision.”LinkedIn · Feb 2026
Kai has the unique talent of knowing technology in depth, while understanding business priorities …
“I have worked with Kai for many years, even before he became my customer (and partner). Kai has the unique talent of knowing technology in depth, while understanding business priorities, building communities and being a human that everyone wants to work with. This is a unique and rare combination in the tech world. Kai is a great leader to work with for IoT, AI and across verticals. He is curious and learning fast by being hands on. Hope to continue working with him for more years.”LinkedIn · May 2026
Four stages, each with its typical trap and one highest-leverage next move. The check places you on one.
The conversation has started, but no one owns a number yet.
Next move: name one business-owned problem with a number.
AI is paying off in a few workflows, but it only assists.
Next move: from assistance to delegated action (one task, an approval gate).
Agents run real, multi-step processes end to end.
Next move: build the reliability layer (evals, guardrails, escalation).
The business is being designed around agents.
Next move: redesign a core value-chain process with agents as the default.
Before money changes hands, both sides know how the engagement will be measured.
You describe the situation; I ask the awkward questions.
Feasibility, timing, cost envelope — in writing. If the assessment does not hold up, the engagement ends here.
Exit here: no invoice
If you prefer, start with a clearly scoped fixed-fee audit instead of a long-running mandate — agentic operating model, architecture, or go-to-market.
Every piece read against the primary source. Where the coverage gets it wrong, the note says so.

Whether you have a concrete IoT or AI mandate in mind or want a strategic read on where you stand, the introductory call is confidential, with no obligation.