Adopting AI in processes
Joint sessions, agents for specific processes, team training. I work with AI agents in my own product every day - and transfer what really works.
Discuss a projectHow clarity comes
The work format appears after the task is analyzed, and not before it.
Finding a real problem
Not “implement AI”, but a specific place where hours or money are currently being lost.
Looking at what already exists
Often a task is completed with a ready-made tool. It's cheaper than any development.
Cutting off the excess
What remains is what really requires a solution.
Assembling and checking
The first verifiable result, not a presentation about the future.
Joint session
We analyze the process, assemble the first solution live and record what’s next. Three o'clock.
MORE ABOUT ASSEMBLY →Team training
Common language and overall picture: what models can do, where they break and how not to give them too much. Experience in teaching and acceleration programs.
TRAINING PROGRAMS →On topic
ALL MEDIA →AirPods 5: Noise-cancelling open fit for $129
Apple claims minus 50% noise versus AirPods 4 with ANC, five hours of operation and Live Translation across the entire line. Russian is not in the list of Apple Intelligence languages, which means that the translation paired with Russian will not work out of the box.
Claude 5: what has really changed for those who build on AI
I'm analyzing a new release without hype: what has become noticeably better in real tasks, and what is just marketing.
Copperhead designs boards in KiCad and commits by stages
The open agent breaks the development of the board into eight stages, each must leave the artifact on disk and go through the ERC/DRC of KiCad itself, otherwise there will be no commit. The gate scheme is transferred to any agent pipeline, where there is something to check the result.