On May 21, 2026, the third day of the webinar "AI.Prom" took place. The first two days focused on AI in sales and manufacturing. The third day was about lawyers, consultants, employee adaptation, and digital twins in VR. Here’s a summary.
AI Lawyer: Contract Review in 5 Minutes
Host Andrey Melkov showcased a product that started as an internal tool and is now used in practice. The main goal is to relieve lawyers from routine tasks: contract review, risk identification, version comparison, and claim preparation.
In the demo, a contract was uploaded. The AI highlighted the parties, identified risks, and provided recommendations — with references to the Civil Code and arbitration practice. For each risk, it suggested a safe wording. There is a mode for comparing two versions of a contract: it shows significant changes (for example, the advance payment period increased from 3 to 5 days) and cosmetic edits. A claim can be generated in less than a minute — the system pulls details from the contract and generates the text.
An important point regarding security: contracts cannot be uploaded to ChatGPT or other cloud services — this violates 152-FZ and leads to the leakage of commercial secrets. The solution works only in the customer's local environment (on your hardware, without data transmission outside). If you need links to recent laws and arbitration practices — you will have to connect regulatory systems like Consultant+ or update the database manually.
AI Consultant: Knowledge Base Always at Hand
Problem: there are numerous regulations, instructions, and memos at the enterprise. A new employee does not know where to find information, distracting mentors. The solution is an AI Consultant that answers questions based on your knowledge base.
Andrey demonstrated this using a regulation on equipment. He uploaded the document and asked, "Who is responsible for replacing fuses after an accident?" — the AI instantly provided the answer from the regulation. The system can also search inventory (stocks, prices), accounting systems, and generate reports. Responses come in 20–30 seconds. It operates 24/7.
HR Platform by Reshat Rustamov: Onboarding on Autopilot
Reshat discussed the system for employee adaptation and training. It addresses the pain point: mentors spend time on repetitive questions, training plans become outdated, and there is no unified control.
How AI works in this platform:
- Newcomer Assessment. The AI conducts interviews (text or voice) to identify knowledge gaps.
- Training Track Formation. Based on the interview and knowledge base, the AI builds an individual plan — which articles, instructions, and exams need to be completed.
- 24/7 Tutor. Employees can ask questions to the system at any time without bothering their manager.
- Quality Control. For the sales department — call analysis: politeness, handling objections, script adherence. The AI transcribes conversations and provides evaluations.
- Continuous Learning. Based on real results (calls, documents), the AI adjusts the track — providing materials on weak points.
All of this is cyclical: analysis → new track → training → analysis again. The system does not forget about the versioning of regulations.
VR and Digital Twins by Alexey Zhukevich
Alexey presented solutions based on virtual and augmented reality. Experience — over 300 cases in 10 years.
Technology levels:
- Digital Twins — a digitized 3D space of the workshop with processes. You can model weather conditions, accidents, without touching real equipment.
- Mixed Reality — a camera looks at the worker's hands and provides hints directly on the part. Ideal for newcomers.
- Classic VR — headset, gloves, full immersion. A person trains on a virtual machine, without risking equipment and life.
Results from cases: learning speed increases by 4 times, engagement — 400%, budget savings on training up to 80%. Especially effective for hazardous industries (chemistry, welding, high-precision equipment). A pilot can be conducted for 400–500 thousand rubles, measure the effect, and then scale up.
Webinar Summary
AI in manufacturing is not fiction but working tools. Lawyers, consultants, HR platforms, and VR simulators are already implemented in real enterprises.
The main thing is data security. No cloud-based ChatGPT. Only a local environment; otherwise, there is a risk of losing commercial secrets and violating the law.
AI does not replace people. It takes on routine tasks: document checks, answers to standard questions, training control. Decisions are made by humans.
Continuity. Learning should be cyclical — after starting work, employees continue to learn based on real mistakes and successes.
If you work with contracts, regulations, or train personnel — I recommend watching the recording to see how it works in live demos. The effect is evident.