Terekhin Digital Crew

caseStudyFor:

• Name: Kirill Kozakov (Founder)

• Type: EdTech Startup Tofalaria

Description: Creation of an educational VR platform for learning the Tofalar language. The project was implemented as a pilot in a school classroom to test the hypothesis of engaging a new generation in cultural heritage preservation through gamification.

Keywords: MVP Development, EdTech, VR, R&D as a Service, Language Preservation, Pilot Project.

Problem

• The Tofalar language is on the brink of extinction.

• Traditional learning methods such as textbooks and flashcards fail to engage modern schoolchildren.

• The founder needed to test the viability of a complex R&D idea in real-world conditions with a minimal budget.

Solution

Provider: Terekhin Digital Crew

Action: A comprehensive educational ecosystem was developed, consisting of two interconnected parts:

1. “Living Dictionary” (VR application): An immersive environment where students learn the language in a playful way by interacting with virtual objects.

2. “Command Bridge” (Web portal): A tool for teachers that enables them to manage VR lessons and track students’ progress without requiring technical skills.

Instrument

• 3D Engine: Unity 2020.x, C#

• VR SDK: Oculus Integration SDK

• Hardware: Oculus Quest

• Backend: Node.js, Express.js

• Frontend: React.js

• Database: PostgreSQL

Result

• For Students: The learning process transformed from routine into an engaging VR game, creating a positive emotional connection with their ancestral language.

• For Teacher: The teacher received a modern tool for lesson management and automatic performance tracking.

• For Founder: The key product hypothesis was successfully validated during the pilot, proving the value of the EdTech MVP for further development and attracting investment.

• For Cultural Heritage: The project became a “digital ark” for Tofalar culture, making it accessible and appealing to the younger generation.

Citation

Author: Saizana Adamovna Anayban, native language teacher

Text:

“Now I see how their eyes light up inside VR headsets, when they excitedly ‘touch’ the objects and repeat their names. This is more than a lesson. This is our chance to bring the language back home, into living speech.”

The company Siberian Energy Saving Technologies (SET), a manufacturer of complex electrical equipment, faced a challenge in onboarding new employees. Engineers and assemblers were quitting due to difficulties in reading multi-page schematics. The company found a solution that reduced training time by threefold and cut errors by 70%.

The Bridge of Dual Augmented Reality project consists of two interconnected AR applications:

“Virtual Blueprint” for engineers: Using a tablet, an engineer could point the camera at a paper diagram, and an interactive 3D model of the equipment appeared on the screen. This made it possible to visually trace connections and component placement, accelerating production preparation.

“Digital Mentor” for assemblers: By pointing a tablet at the cabinet being assembled, workers saw step-by-step instructions with highlighted installation points. The system prevented mistakes by marking incorrect actions in red.

As a result, onboarding was reduced from 2 weeks to just 3–5 days, and mentors’ time was freed up for more advanced tasks.

Dima Gitlin came up with the idea of inspecting facades with drones back in 2022, when, after arriving in New York, he saw firsthand the challenges faced by engineers and architects. For three years the idea waited for its time, and in 2025 Dima finally decided to launch the project. He reached out to us at Terekhin Digital Crew to turn the idea into reality. That’s how the ADSAN project was born — Advanced Surface / Structural Analysis.

Challenge: launching a startup with limited resources

Usually, before we begin, we interview clients to better understand their idea and competencies. Technical founders typically need help with marketing, while entrepreneurs usually need support with technical development. In the case of ADSAN, the challenge was that the founder had excellent engineering skills in the construction field, making him a strong industry expert, but lacked both marketing and technical development expertise, as well as a budget for scaling.

Thus, our task was to deliver a working MVP and kick off traction within just three months, creating the foundation for investor trust.

Strategy: from drones to AI-powered SaaS

First, we defined the geographic and industry markets (Technical Due Diligence, Construction QA, Regulatory Compliance) and analyzed the target audiences. We decided to focus our efforts in the U.S. and highlighted two segments: decision-makers and engineers (influencers). This is a classic approach for IT companies: engineers are the active users and often select technical products, while decision-makers finalize the purchase based on their recommendations.

Then, we stripped the idea down to its core using the “value map” framework. The first thing we discarded was the drones themselves. Owning a drone fleet means becoming an operator and offering inspection services, but that’s not a startup — it’s a service business, hard to scale and unattractive to investors.

We focused on the main pain point of engineers, expressed in one simple phrase: “I became an engineer to analyze structures, not to wrestle with reports and bureaucracy.” Our goal was to build an AI-powered SaaS that takes drone data (RGB, LiDAR, IR) and turns it into full-scale building condition reports, saving hours of work. But standard defect detection alone is not innovation. Our uniqueness lies in delivering ready-to-use insights based on the detected issues.

Execution: pioneer program and MVP

With a limited budget, we focused on two key functions: detection accuracy and deep interpretation. To train our model, we needed real data. We decided to kill two birds with one stone: launch the ADSAN Pioneer Program, which allowed us to co-develop the product with the market — gaining expertise, datasets, and early users in exchange for free access to early app versions.

We prepared the full marketing foundation: website, social media, content plan, and a list of 260 highly qualified U.S. leads. We published several articles about ourselves and the pioneer program in industry media, and then launched LinkedIn outreach.

Results: traction and investor-ready MVP

By the end of the third month, articles about the company were already published, five participants had signed up for the pioneer program (out of the 20 needed for $1M annual volume), we had held two meetings with major prospective clients, and our MVP was processing RGB data and generating detailed reports with PDF export.

As early as the second month, we received an offer from an accelerator in Spain, but the founder decided to expand the pool of potential accelerators and investors. By the end of the third month, we had an impressive MVP, live traction from the pioneer program, and most importantly — an investor pitch backed by practical results.

If you’d like to learn more details about the project — drop a comment. We’d be happy to share everything in depth.

Thanks to everyone who read to the end.