System root
Verified workflows // 2026

AI agent and business automation cases

Each workflow is tied to a working project, its documentation or a public write-up. We show the input, system actions, failures and the point at which a person remains responsible.

On Geron Labs

Client cases and projects

EST.EPIL: AI receptionist

How Geron Labs connected EST.EPIL's website and Telegram to YCLIENTS for current prices, available slots, confirmed bookings and human handoff.

YCLIENTSTelegramwebsite widgetbookings
View case

REGION SNAB: proposal agent

A workspace for tender extraction, product matching, internal and web search, specification checks, calculations and Excel/PDF proposal drafts.

tendersproduct searchExcel / PDFFastAPI
View case

LISVENT: requests and 1C

File exchange, semantic retrieval, model-number checks and manager review connect free-form customer requests to 1C product records.

1Cproduct catalogExcelsemantic search
View case

DOLOROSA: Avito and Telegram hub

Connecting Avito, MoySklad and Telegram for item-specific answers, product photos, order events and a seamless manager handoff.

AvitoMoySkladTelegram hubresale
View case

RDK: requests and access delivery

Email, Telegram and MAX requests, validation, M4 objects and equipment, staff approval, access instructions, SMS delivery and GSM relay commands.

M4Telegram / MAXvideo surveillanceSMS
View case

Telegram Mini App for fitness coaches

How Fit Bot grew from a Google Sheets MVP into an application with coach and client roles, FastAPI, SQLAlchemy and Telegram signature validation.

Telegram Mini AppFastAPIfitnessHMAC
View case

Website with a controlled 3D model

How a ready-made astronaut GLB became part of the interface through custom poses, trajectory controls and smooth Three.js animation.

Three.jsGLBGoogle AI Studio3D animation
View case
Search pipeline

How an AI agent retrieves and matches data

A single language-model prompt is not enough for 1C catalogs or tender work. Retrieval is a staged process, and the source and reason for each match remain visible.

01

Exact match

Start with the SKU, model, identifier and normalized name. A reliable exact result should not be replaced by a model guess.

02

Semantic retrieval

When wording differs, embeddings retrieve semantically close items. They produce candidates, not a final answer.

03

Specification checks

Code and the model compare power, size, material, package contents and other required fields. Any mismatch remains visible.

04

External search

The web is used only when needed. The result keeps its source URL, while price and product fit remain subject to human review.

Case index

Other projects and write-ups

The status distinguishes a public write-up from a case that is still being reviewed. A draft is not presented as a published result.

Article ready

Product photo processing automation

Product photos are cleaned, assembled into listing covers and reviewed before publication. One verified batch contained 1,594 photos for 280 products.

Public processing cost: about $0.10–0.15 per photo. Output quality depends on the source image and is reviewed by a person.

product photosmarketplaceslisting covershuman review
Public case

Field staff dispatch and photo reports

A dispatcher assigns jobs, field workers update statuses through Telegram, and photos are stored separately from operating spreadsheets. The interface highlights incomplete and problematic visits.

Time savings were not measured. The public write-up focuses on the workflow and its platform trade-offs.

field staffTelegramphoto reportsdispatch
Read on vc.ru (RU)

Where to start

Choose one repeated workflow and bring real input files or conversations. We identify the data source, acceptable error, agent actions and the point that requires human confirmation.

Review a workflow
SYS: COOKIE

We use technical browser storage so the site works correctly. Privacy Policy