Manufacturing strategy with AI
I identify where a company loses time and management attention, which decisions repeat, and where automation is appropriate. The result is a sequence of implementation steps with clear responsibility boundaries.
Rinat Sultanov
I implement AI in manufacturing companies in practice
I implement AI across sales, data and management inside an operating food manufacturing business. Solutions that prove useful in practice are then adapted for other manufacturers.
Director of Development at Kazan Delicacies. I work at the intersection of commercial operations, data and automation.
- Agent-led salesresearch, qualification and first contact
- Data & management1C, analytics and management visibility
- Productionvideo monitoring, accounting and automation
- VisibilitySEO + AI search, structured content
For me, AI implementation is a combination of processes, data and clear responsibility between people and systems.
I test solutions inside an operating manufacturing business first, then adapt the approach for other companies.
systems
What is already working and what is in progress
The list is updated from this server. Client names and sensitive internal details are not published. System descriptions are currently in Russian.
connecting…
08finished
10in progress
Now
Finished
01 / work
What I do
Analytics with AI
I connect data from 1C, spreadsheets and other sources so management can see the current picture and deviations without waiting for a manually assembled report.
Faster sales with agents
An agent handles repeatable first-touch work: company research, qualification, first-contact preparation and a manager brief. People take over when negotiation, terms and trust matter.
SEO + AIO / GEO
I structure websites so search engines can find them and AI assistants can interpret them consistently. The same content layer should make sense to both people and machines.
02 / method
Company design first, tools second
-
01
Constraints and responsibility
Quality, safety, client requirements and industry rules define the boundaries of automation before a technology is selected.
-
02
The current process
We document how work happens today: who decides, where data comes from, where waiting and manual transfer occur, and which actions repeat.
-
03
Connected data
Accounting systems, documents and other sources are connected so information is available without constant manual copying.
-
04
Reversible tasks go to the system
AI starts with tasks that can be checked, stopped or rolled back. Irreversible decisions remain under human control.
Each stage should produce a useful result on its own: a clearer process, connected data, time savings or a measurable improvement in quality.
03 / in practice
What is already working inside an operating company
A site for search and AI discovery
The catalog and pages are structured so people, search engines and AI assistants can read the same information consistently.
First-touch sales work
An agent finds and qualifies companies, prepares the first contact and passes context to a manager for the next stage. Case with numbers →
Production control
Video and operational data are used to observe shifts and events on the production floor, with deeper automation introduced gradually.
04 / about
Twenty years in commerce and manufacturing.
Now — development through AI.
My experience combines commercial operations, food manufacturing and digital systems. I therefore treat AI as part of the operating model rather than a separate IT function.
I am Director of Development at Kazan Delicacies, a halal meat-processing business in Kazan. Revenue grew from RUB 58.6 million in 2023 to RUB 869 million in 2025. I do not attribute that growth to AI; the value for this work is having a live business where digital solutions can be tested against real processes.
Inside the company I develop agent-led sales, 1C data workflows, video monitoring, SEO and AI-search visibility. This makes it possible to evaluate solutions by how they perform in day-to-day operations, not only by vendor presentations.
- 2001First business: building websites while at university
- 2001–06Astrakhan State Technical University — Applied Informatics in Services, PR specialization. Informatics Analyst
- 2006–18Meat industry: turnkey retail, distribution to 1,000+ accounts, commercial leadership at a plant, federal retail chains
- 2018–22Co-founder of a cleaning-service company: Astrakhan, Krasnodar, Dubai
- 2020–22Master’s degree. Management. Enterprise management — Estonian Entrepreneurship University. MBA
- 2022 →Kazan Delicacies, Director of Development: ×14.8 revenue in two years, 64 SKUs
- 2023–26Digital operations layer: agent-led sales, 1C, video monitoring, AI workflows, SEO + GEO
05 / perspective
AI in manufacturing starts with processes and data
First define constraints and responsibility, then document the process and connect the data. From there, individual tasks can be transferred to systems and automation can deepen gradually.
Starting with a tool makes it easy to end up with isolated pilots. I prefer the reverse order: business structure and success criteria first, technology second.
06 / formats
Three ways to start
Diagnosis
For a leader who wants to identify where AI can create practical value and in what sequence to proceed.
In 2–3 weeks I review processes, data, tools already in use and vendor proposals.
The output is a map of opportunities, constraints and implementation sequence.
90-day pilot
One measurable process: sales, analytics, inbound requests, the website or production control.
We define the baseline and success criteria before launch, then compare them with actual changes after 90 days.
A working system and clear economics for the next step.
Change architecture
For companies that need a sequence of changes rather than a single pilot.
I work on the owner’s side: sequencing the work, setting result criteria, and coordinating internal and external teams.
No more than 2–3 companies at once.
07 / who I work with
Who I work with
Mid-size and large manufacturers and FMCG companies that need to connect commercial operations, data, management and automation.
Typically these are businesses with established processes and teams where further growth requires more systematic management.
- growth is increasing the amount of manual work;
- data exists but is collected and interpreted inconsistently;
- AI tools have been tried but have not yet become a coherent system;
- automation, scaling or entry into new markets is ahead.
When my format is not a fit
- You only need a one-off setup of a single service or tool.
- There is no willingness to change the underlying process or responsibility model.
- There is no process owner, usable data or baseline to measure.
- For micro-businesses my format is usually excessive; it is designed for existing processes, data and teams.
If the company has established processes and a clear business problem, one measurable workflow is enough to start.
08 / technical transparency
How search engines and AI agents see this site
Below are saved results from external checks and links to the same tools for any other website.
PageSpeed Insights
- 100Performance
- 100Accessibility
- 100Best Practices
- 100SEO
agent browsing 3/3 · FCP 0.2 s · LCP 0.3 s · TBT 0 ms · CLS 0 · SI 0.8 s
Open the PageSpeed report ↗Is Your Site Agent-Ready?
100 Level 5 · Agent-Native
Open the scan ↗Run the same tools on your site
09 / external references