Alexandr Chibilyaev reveals AACFlow enterprise connectors for 1C, Diadoc, EGAIS, SMEV3, FNS, Chestny Znak, and the Kontur ecosystem — AI agents operating inside the regulated Russian economy.
There's an entire economy that AI platforms ignore. It's not small. It's not niche. It's the backbone of every country: accounting, tax filing, electronic document exchange, alcohol tracking, product labeling, government service integration.
In Russia, this economy runs on infrastructure that no Western AI platform has ever heard of: 1C, Diadoc, EGAIS, SMEV3, FNS, Chestny Znak, the Kontur ecosystem. These aren't "integrations" you can add with a Zapier connector. They're complex, regulated, compliance-heavy systems that require deep domain knowledge to interface with correctly.
AACFlow connects to all of them. Not with stubs. Not with placeholder configurations. With real connectors that understand the data models, the authentication protocols, the compliance requirements, and the failure modes of each system.
This is what makes AACFlow the only AI agent platform that bridges the gap between modern AI and the regulated economy.
If you work in Russian business, you don't need me to explain 1C. For everyone else: 1C:Enterprise is the ERP/accounting platform used by virtually every company operating in Russia — from sole proprietors to the largest enterprises. It handles accounting, tax, payroll, inventory, CRM, document management, and industry-specific operations.
1C is not a single API. It's a platform with multiple integration surfaces:
OData (REST API) — the modern interface. Exposes 1C data objects (documents, catalogs, registers) via a standardized REST protocol. Our connector supports OData with automatic metadata discovery — it reads the 1C configuration schema and adapts its queries accordingly.
HTTP Services — custom REST endpoints published by 1C configurations. Each 1C deployment is unique — companies heavily customize their configurations. The connector allows configuring custom HTTP service endpoints with parameter mapping.
File Exchange — many 1C integrations still use file-based exchange (XML, Excel, CSV via shared directories). The connector supports watching file directories and parsing standard 1C exchange formats.
The connector extracts documents into the knowledge base with structured tag mapping:
An AI agent connected to 1C can answer questions like: "What invoices are overdue by more than 30 days?", "Show me the P&L by department for Q1", "Which counterparties have outstanding debts above 100,000₽?" — not by guessing, but by querying actual accounting data.
Diadoc (by SKB Kontur) is the dominant electronic document exchange (EDO/EDI) platform in Russia. It handles legally significant document flow: invoices, acts, contracts, waybills — all digitally signed, all legally binding, all exchanged between companies without paper.
The regulatory stakes are high. An electronic invoice must be:
Signed with a qualified electronic signature (KEP)
Formatted according to FNS-mandated XML schemas
Transmitted through an accredited EDO operator
Archived for the legally required retention period
Available for tax authority inspection on demand
Our Diadoc connector handles:
Incoming documents — invoices, acts, UPD (universal transfer documents), contracts received from counterparties
Outgoing documents — documents sent for signing, with status tracking (sent, delivered, signed, refused)
Document content extraction — PDF and XML documents are parsed, text is extracted for embedding, and structured fields (amounts, dates, counterparties, tax rates) are mapped to tags
Signature verification — the connector validates that documents have valid electronic signatures before accepting them
An AI agent connected to Diadoc can: check if a specific invoice has been signed by the counterparty, alert on documents nearing signature deadlines, reconcile incoming invoices against purchase orders in 1C, and flag discrepancies between document amounts and actual deliveries.
EGAIS (Unified State Automated Information System) is Russia's alcohol production and distribution tracking system. Every bottle of alcohol produced, imported, or sold in Russia must be tracked in EGAIS — from the distillery to the retail shelf. This is regulated by Rosalkogoltabakcontrol and carries criminal liability for non-compliance.
The EGAIS connector interfaces with:
Waybills (TTN) — incoming alcohol shipments are registered, verified against the physical delivery, and confirmed or rejected
Production records — producers report every batch produced with volume, strength, and composition
Retail sales — every retail sale of alcohol must be recorded in EGAIS (via the cash register integration)
Inventory reconciliation — periodic reconciliation of actual stock against EGAIS-registered volumes
Marking codes — integration with the federal special / excise stamp system
For a restaurant chain using iiko (restaurant management) + EGAIS, an AACFlow agent can: automatically confirm incoming alcohol waybills by matching EGAIS electronic TTNs against physical delivery records from iiko, alert on discrepancies before the confirmation deadline, generate mandatory alcohol declaration reports, and flag inventory that's approaching its EGAIS confirmation deadline.
The compliance window for EGAIS confirmations is tight — 24 hours for urban retail, 72 hours for rural. Missing a confirmation means the alcohol is legally unsellable. An always-on agent that monitors EGAIS waybills and alerts 2 hours before deadline eliminates this risk.
SMEV3 (System of Interdepartmental Electronic Interaction, version 3) is Russia's government service bus. It connects federal, regional, and municipal government information systems. Through SMEV3, organizations can query government registries, submit applications, and receive official responses — all electronically.
The SMEV3 connector supports:
Service discovery — querying the SMEV service registry to identify available government services
Request submission — formatting and sending requests according to SMEV3 message specifications (SOAP/XML with GOST cryptographic signing)
Response processing — extracting structured data from government responses and normalizing it for the knowledge base
GOST cryptography — integration with CryptoPro CSP for electronic signature and encryption
Use cases:
Query the Unified State Register of Legal Entities (EGRUL) before signing a contract with a new counterparty
Submit statistical reports to Rosstat
Verify property rights via Rosreestr
Check an individual's tax status via FNS
Submit applications to government agencies with electronic signatures
For a legal department, an AACFlow agent connected to SMEV3 can: automatically verify every new counterparty against EGRUL, FNS disqualified persons registry, and the Federal Bailiff Service (FSSP) debtor database — flagging any red flags before a contract is signed.
The Federal Tax Service (FNS) provides APIs for taxpayer data access. The FNS connector integrates with:
EGRUL/EGRIP — legal entity and individual entrepreneur registries
Tax debt verification — checking an organization's or individual's tax arrears
Self-employed status verification — validating that a contractor is registered as self-employed (NPD) with valid status
VAT invoice verification — checking that an incoming invoice is registered in the FNS system
For an accounting firm processing hundreds of counterparties monthly, an agent can: verify every counterparty's tax status before processing payments, flag counterparties with tax debts (indicating financial instability), and validate self-employed contractors to ensure proper tax treatment of their invoices.
Chestny Znak (Honest Mark) is Russia's national product labeling and traceability system. It covers an expanding list of product categories: footwear, textiles, perfumes, tires, cameras, dairy products, packaged water, beer, dietary supplements, antiseptics, medical devices, and more.
Each product unit carries a unique Data Matrix code. Every movement through the supply chain — production, import, wholesale, retail sale — is recorded in Chestny Znak. Selling unlabeled products or failing to report sales carries severe penalties.
The Chestny Znak connector handles:
Code emission — ordering labeling codes for production
Shipment tracking — recording product movements through the supply chain
Retail sales registration — reporting sales to Chestny Znak (direct integration or via cash register OFD)
Inventory reconciliation — matching physical stock against Chestny Znak-registered volumes
Compliance monitoring — checking that all products in inventory have valid, active labeling codes
An AACFlow agent can: automate the code ordering process when inventory drops below threshold, alert on products with labeling codes approaching expiration, reconcile physical inventory (from 1C) against Chestny Znak data, and generate compliance reports for inspections.
SKB Kontur is Russia's largest B2B SaaS company, providing a suite of products that form the digital backbone of business operations. AACFlow connects to the full ecosystem:
Kontur.Accounting / Kontur.Buhgalteria — cloud accounting with tax filing integration. The connector extracts financial documents: invoices, acts, waybills, bank statements, tax declarations.
Kontur.Elba — simplified accounting for sole proprietors and micro-businesses. Extracts revenue, expenses, tax calculations, and reporting status.
Kontur.Extern — electronic tax reporting. Submit tax declarations, receive FNS responses, track filing status. An agent can monitor filing deadlines, verify declaration acceptance, and alert on FNS inquiries.
Kontur.Diadoc — EDO (electronic document exchange), the same Diadoc product described above, accessed through the Kontur platform.
Kontur.Focus — counterparty verification. Check any Russian legal entity or entrepreneur: registration data, financial statements, arbitration cases, government contracts, affiliated entities, negative signals.
Kontur.Sign — electronic signature service. Issue, manage, and use qualified electronic signatures (KEP) for signing documents and submitting reports.
An enterprise running on the Kontur ecosystem can connect all of these to AACFlow. An AI agent gains visibility into every financial and legal aspect of the business — and can act on it.
Government and regulated-industry data carries compliance obligations. Our architecture respects them:
Data residency. When self-hosted, all data stays in the customer's infrastructure. When using AACFlow Cloud, data is stored in geographically appropriate regions. Government data never traverses jurisdictions it shouldn't.
Encryption at rest. API keys, credentials, and electronic signature keys are stored encrypted. The encryption key is either AACFlow-managed (cloud) or customer-managed (self-hosted).
Audit logging. Every agent action involving government data is logged with immutable records: what data was accessed, by which agent, at what time, for what purpose. These logs are exportable for compliance audits.
No training on customer data. Customer data — government or otherwise — is never used for model training. Agent interactions with government APIs are deterministic and auditable, not "model hallucinations about tax law."
Human-in-the-loop. Agents operating on compliance-sensitive data can be configured to require human approval for irreversible actions (submitting a tax declaration, confirming an EGAIS waybill, signing a contract). The agent prepares the action, explains its reasoning, and waits for approval.
Most AI platforms target the unregulated, low-stakes parts of business: marketing copy, chatbots, internal knowledge bases. That's valuable work, but it's only half the picture.
The other half is the regulated, compliance-heavy, infrastructure-dependent work that actually keeps businesses running: accounting, tax, legal, logistics, government reporting. This work is more complex, carries higher stakes, and — critically — has almost no AI automation available today.
AACFlow connects AI agents to this world. Not as an afterthought. As a first-class design goal.
A 1C accountant who spends 40% of their time on repetitive data entry can be augmented by an agent that pulls bank statements, matches them to invoices, and flags discrepancies. A compliance officer who manually checks every counterparty against five different registries can have an agent do it automatically for every new contract. A logistics manager who reconciles Chestny Znak codes against physical inventory can have an agent run the reconciliation daily and report exceptions.
This isn't "AI will replace jobs." This is "AI will handle the repetitive, compliance-driven, data-intensive work that humans are bad at — so humans can focus on judgment, strategy, and relationships."
There is no other AI agent platform that connects to 1C, Diadoc, EGAIS, SMEV3, FNS, Chestny Znak, and the Kontur ecosystem. Western platforms don't know these systems exist. Russian AI companies are focused on LLMs, not agent orchestration platforms.
AACFlow is unique. We've invested the engineering effort to build real connectors to regulated infrastructure — not just list names on a marketing page. Each connector implements the ConnectorConfig contract. Each handles authentication, pagination, error recovery, and data normalization. Each is tested against real APIs with real data.
For businesses operating in Russia's regulated economy, there is no alternative. AACFlow isn't just the best choice for connecting AI agents to government infrastructure. It's the only choice.
