Shalom Arhebamen
Chief Technology Officer
Node JS
Go
JavaScript
Python
TypeScript
Data Pipeline & Feature Engineering: 4/10
Active 5 days ago
+7 (950) 8701790 Invite to interview
Message
Download CVCV
Overview
Technical skills
Timeline
Roles
Overview
SDET / Test engineer (senior-level) focused on building robust automated test suites and reliable API services. The strongest proven skill is end-to-end and integration testing backed by Playwright config and E2E tests (next-bun-threaded-chat/playwright.config.ts and e2e/*.e2e.ts) plus FastAPI integration/unit tests (deribit-index-price-history-api/tests/test_prices_api.py). There is limited evidence of formal performance/load testing or published bug-tracker driven root-cause analyses in the public code.
Technical skills
Node JS
Go
JavaScript• Junior
Python• Senior
TypeScript• Senior
Node JS
Prisma
Puppeteer
BullMQ
Python
Alembic
Celery
Asyncio
Aiohttp
SQLAlchemy
FastAPI
Pydantic
aiogram
Databases
SQLite
MinIO
PostgreSQL
Redis
AI/ML
OpenRouter
LLM
NLP
RAG
DevOps
Grafana
Loki
OpenTelemetry
Prometheus
Docker Compose
Rest API
CI/CD
Docker
GitHub Actions
Helm
Kubernetes
Traefik
QA
Vitest
Playwright
Selenium
Frontend
Next.js
React.js
Docusaurus
Vue.js
Timeline
Chief Technology Officer (CTO)
•
Executive
Worken
•
Full-Time
Led the technical strategy, system architecture, and reliability for Worken AI, staying hands-on with engineering. Developed the AI agent platform components, including agent/workflow runtime, managed MCP, knowledge/RAG, model catalog, billing, and internal APIs. Delivered and stabilized integrations such as Bitrix24 and Telegram, improved security with scoped authorization and safer mutation patterns, and drove production delivery using Docker, Kubernetes, Traefik, and GitHub Actions. Conducted architecture/code reviews, regression testing, and incident response across backend, frontend, and infrastructure.
Bitrix24since 2026
GitHub Actionssince 2026
CI/CD
Dockersince 2026
Kubernetessince 2026
Traefiksince 2026
RAGsince 2026
Chief Technology Officer (CTO) / Technical Director
•
Executive
Worken
•
Full-Time
Serves as CTO after prior backend/platform engineering work, owning AI platform architecture, technical strategy, and production reliability while staying hands-on. Leads development of AI agent/workflow runtime, managed MCP, knowledge/RAG, model catalog, billing, and related integrations/internal APIs. Delivered and stabilized integration pilots for Bitrix24 and Telegram bots, improving platform security via scoped authorization and safer mutation handling. Owns production delivery using Docker, Kubernetes, Traefik, and GitHub Actions, and performs architecture/code reviews and incident response.
Bitrix24
RAG
GitHub Actions
Docker
Kubernetes
Traefik
Full-stack TypeScript Developer (contract)
•
Middle
Confidential
•
Contractor
Built full-stack functionality for a B2B procurement and project management platform using Vue 3 and TypeScript-based services. Implemented an end-to-end workflow from procurement to projects, including budgets/estimates persistence and asynchronous orchestration with BullMQ over PostgreSQL. Designed revision-aware two-way synchronization for documents and created modules for approvals, versioning, and change logs, while expanding RBAC to platform roles and capabilities. Added regression coverage for APIs, queues, repositories, migrations, and Vue components, and worked with Docker, Kubernetes, Helm, MinIO/S3, and Redis.
TypeScriptsince 2026
Vue.jssince 2026
PostgreSQLsince 2026
BullMQsince 2026
Redissince 2026
MinIOsince 2026
Docker
Kubernetes
Helmsince 2026
Full-stack TypeScript Developer (contract)
•
Middle
Confidential B2B project
•
Contractor
Worked on a short-term B2B procurement and project management platform development using TypeScript and Vue 3. Implemented end-to-end procurement-to-project workflows including project creation, estimates/budgets, and asynchronous orchestration with BullMQ backed by PostgreSQL persistence. Built revision-aware bidirectional synchronization with idempotency and stale-update protection. Extended RBAC for independent platform roles and capabilities and added regression test coverage across API, queues, repositories, migrations, and Vue components.
TypeScript
Vue.js
BullMQ
PostgreSQL
Redis
Docker
Kubernetes
Helm
MinIO
Developer Support Engineer / API
•
Middle
Multilogin Software Ltd.
•
Full-Time
Supported developers integrating REST APIs and browser automation scenarios using Selenium/Puppeteer/Playwright and WebDriver. Diagnosed issues in APIs, proxy/networking, browser sessions and automation scripts, producing reproducible examples and troubleshooting guides. Maintained developer-facing documentation using Docusaurus and communicated recurring product/API issues back to engineering teams with technical context. Provided troubleshooting examples using Python and JavaScript.
Rest API
Selenium
Puppeteer
Playwright
Docusaurus
Python
JavaScript
Senior AI/ML Engineer
Confidence: High Generalist
Conversational backend engineer (senior-level) building deterministic NLQ-to-SQL pipelines and Telegram bot integrations. The strongest proven skill is designing a robust NLQ parsing and deterministic SQL builder with comprehensive unit and integration tests, as evidenced by src/intent/rules_parser.py and src/sql/builder.py plus test suites under tests/. The public code does not show model training, model-serving infrastructure, quantization or GPU optimization work, nor broad MLOps pipelines for deployed models.
Model Architecture & Training
1/10
How well models are designed and trained
Minimal model-building evidence; the code implements a deterministic rules-based NLQ parser and an optional LLM JSON parser call rather than any model architecture or training loop.
Data Pipeline & Feature Engineering
4/10
How data is prepared for models
Clear, pragmatic data ingestion and schema handling with batching, transactional inserts, and migration scaffolding for Postgres.
Experimentation & Evaluation
3/10
How results are measured and tested
Good unit and integration test coverage for the NLQ->SQL pipeline and database integration; tests validate correctness and guard against regressions.
MLOps & Deployment
1/10
How models are shipped to production
Basic operational scripts and DB lifecycle utilities exist but there is little evidence of production MLOps or model-serving infrastructure.
Computational Efficiency
3/10
How efficiently computing resources are used
Some attention to efficiency via batching and async connection pooling; no GPU, quantization, or heavy compute optimizations are present.
Research Depth & Innovation
1/10
Depth of research and new ideas
No implementation of novel research, custom layers, or reproduced paper results; the work is engineering-focused and deterministic rather than research-forward.
Expertise
Conversational AI & Chatbots• Middle
LLM• Middle
Industries
Media & Entertainment• Middle
Technologies
Python• Senior
PostgreSQL
MinIO
NLP
LLM
RAG
Asyncio
Pydantic
aiogram
OpenRouter
ElevenLabs• mentioned only
MySQL• mentioned only
Recommendations
- Develop conversational analytics features and NLQ-to-SQL connectors that require strict correctness and SQL parameterization.
- Build backend services that integrate async DB pools, transactional data loading, and isolated-schema integration tests.
- Implement production-grade bot workflows and access-controlled Telegram integrations using aiogram and robust testing.
- Extend deterministic parsers with a carefully instrumented LLM fallback that includes rate limiting, retries, and strict output validation.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Senior Backend Developer
Confidence: High Data Platform
Backend engineer (senior, 4.5) specializing in reliable async data ingestion and API services. The strongest proven skill is building robust async ingestion and DB integration, demonstrated by app/services/price_service.py together with app/db/repository.py and alembic migration scripts that implement advisory locks, upsert logic, and session management. There is limited public evidence of large-scale distributed architecture design, production secret management practices, or advanced operational runbooks.
API Design
6/10
How well APIs are designed
API design shows thoughtful request validation, pagination helpers and structured error contracts, but lacks explicit versioning, idempotency key patterns, or documented API lifecycle strategies.
Evidence
arhebs/deribit-index-price-history-api/app/api/pagination.py: build_paginated_response
arhebs/deribit-index-price-history-api/app/api/routes/prices.py: list_prices / latest_price / list_prices_range with validation and HTTPException error payloads
arhebs/next-bun-threaded-chat/src/app/api/chat/route.ts: POST handler with message validation and controlled system prompt construction
Data Layer & Database
7/10
Working with databases
Strong, practical data-layer work with alembic migrations, a repository pattern, upsert tuning and explicit transaction/session management; demonstrates schema evolution and DB-level concurrency control.
Evidence
arhebs/deribit-index-price-history-api/alembic/versions/20260116_1819_create_price_points.py: migration script
arhebs/deribit-index-price-history-api/app/db/repository.py: upsert_price_point (INSERT ... ON CONFLICT), list/count/get queries
arhebs/deribit-index-price-history-api/app/db/session.py and app/db/locks.py: session scope, engine lifecycle and advisory lock usage
Scalability & Performance
6/10
Handling load and speed
Evidence of scalable design via scheduled workers and advisory locks to avoid duplicate ingestion and Celery for decoupling, but no explicit cache-invalidation, load testing, or advanced tuning artifacts.
Evidence
arhebs/deribit-index-price-history-api/app/services/price_service.py: poll_and_store_prices with advisory lock and session lifecycle
arhebs/deribit-index-price-history-api/app/workers/celery_app.py: create_celery_app and beat schedule
arhebs/deribit-index-price-history-api/app/deribit/client.py: HTTP client with timeout and explicit error mapping (rate limit handling)
System Architecture
6/10
Overall system structure
Clear separation of concerns (API, service, repository, worker) and small-service decomposition; system boundaries are pragmatic but there is no evidence of multi-service contracts, cross-team orchestration, or large-scale decomposition trade studies.
Evidence
arhebs/deribit-index-price-history-api/app/api/routes/prices.py and app/services/price_service.py: API -> service -> repository layering
arhebs/deribit-index-price-history-api/app/workers/tasks.py and celery_app.py: worker/service separation
arhebs/next-bun-threaded-chat/: separable frontend and API route modules (chat/tooling) showing multi-process responsibilities
Security & Auth
4/10
Protecting data and access
Input validation and structured error responses are present and client errors are mapped to domain codes; there is limited evidence of auth/token lifecycle, secret rotation, or an audit of dependencies for supply-chain risks in code artifacts analyzed.
Evidence
arhebs/deribit-index-price-history-api/app/api/routes/prices.py: ticker validation with HTTPException payloads
arhebs/deribit-index-price-history-api/app/core/config.py: BaseSettings usage for config
arhebs/deribit-index-price-history-api/app/deribit/client.py: maps HTTP errors and rate limits into typed exceptions
Reliability & Observability
6/10
Stability and monitoring
Good observability and reliability primitives such as structured logging, explicit error mapping, tests and guarded teardown of DB locks; resilience patterns like retries/backoff/circuit-breaker are not prominent in the code reviewed.
Evidence
arhebs/deribit-index-price-history-api/app/core/logging.py: JsonFormatter and configure_logging
arhebs/deribit-index-price-history-api/app/services/price_service.py: try/finally releasing advisory lock and engine.dispose
arhebs/deribit-index-price-history-api/tests/integration/test_db_locks.py and other tests: integration-level tests exercising reliability
Verified artifacts
Expertise
Backend AI & LLM• Middle
Databases & Vector Storage• Senior
Messaging & Real-time• Middle
Python• Senior
Microservices & API Architecture• Middle
Industries
Artificial Intelligence• Middle
Financial Services• Senior
Technologies
Redis
SQLite
Bun• mentioned only
Docker• mentioned only
Persistent• mentioned only
Recommendations
- Build high-throughput ETL/time-series ingestion pipelines (worker + advisory lock + upsert patterns).
- Develop backend APIs for AI-agent tooling that integrate confirmation-gated writes and tool output normalization.
- Implement end-to-end observability and operational playbooks for async workers (backoff, retries with jitter, SLI/SLO hooks).
- Lead design work on schema evolution and migration strategies for multi-tenant or high-ingest databases.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Senior QA Engineer
Confidence: Medium SDET
SDET / Test engineer (senior-level) focused on building robust automated test suites and reliable API services. The strongest proven skill is end-to-end and integration testing backed by Playwright config and E2E tests (next-bun-threaded-chat/playwright.config.ts and e2e/*.e2e.ts) plus FastAPI integration/unit tests (deribit-index-price-history-api/tests/test_prices_api.py). There is limited evidence of formal performance/load testing or published bug-tracker driven root-cause analyses in the public code.
Test Automation Frameworks
6/10
Building automated tests
Solid test automation platform evidence: Playwright E2E with selector discipline, dedicated Playwright config (retries, trace-on-retry, webServer), and unit-level test fixtures and mocks that isolate external services.
Evidence
next-bun-threaded-chat/playwright.config.ts
next-bun-threaded-chat/e2e/chat.e2e.ts
next-bun-threaded-chat/src/app/api/chat/route.test.ts
Test Coverage & Strategy
6/10
What and how to test
Good test strategy with negative-paths and boundary checks present across unit, integration, and E2E; tests encode failure modes, truncation/boundary behavior, confirmation gating, and skip/flagging for flaky AI-dependent scenarios.
Evidence
next-bun-threaded-chat/src/lib/chat/tool-part-normalize.test.ts
arhebs/deribit-index-price-history-api/tests/test_prices_api.py
next-bun-threaded-chat/e2e/real-model.e2e.ts
API & Integration Testing
6/10
Testing how parts work together
API and integration testing shows dependency overrides, monkeypatching, TestClient usage, integration DB tests, and repository-layer tests ensuring correctness of upserts and advisory-lock-protected operations.
Performance & Load Testing
Testing speed under load
Not evidenced in public code
Bug Reporting & Analysis
2/10
Finding and describing bugs
Some in-code error handling and teardown analysis plus tests simulating failure modes exist, but there are no published reproducible bug reports, root-cause issue threads, or dedicated flaky-test investigations in the analyzed files.
Evidence
next-bun-threaded-chat/src/app/api/chat/route.test.ts
tg-group-analytics-bot/apps/bot/src/handlers/analyze.ts
CI Test Integration
5/10
Running tests automatically
CI/test-integration signals exist: Playwright CI-aware retries and trace settings, scripts to run E2E/real-model variants, and a vitest config; however, no CI workflow files (workflows matrix, selective diff runs, quarantine mechanics) were present in the analyzed files.
Evidence
next-bun-threaded-chat/playwright.config.ts
next-bun-threaded-chat/package.json
tg-group-analytics-bot/packages/shared/vitest.config.ts
Expertise
SDET & Test Engineering• Senior
API Testing & Contract Validation• Senior
Industries
Blockchain & Crypto• Middle
Technologies
Go
TypeScript• Senior
Selenium
Playwright
Docker Compose
SQLAlchemy
FastAPI
Celery
Vitest
Aiohttp
Alembic
Docker• mentioned only
Telegraf• mentioned only
Recommendations
- Design and own E2E and integration test infrastructure for web + API stacks, including flaky-test quarantine, artifact collection, and per-test retry strategies (use the existing Playwright setup as a base).
- Implement contract/schema testing and CI gates for API surfaces (extend current FastAPI tests with schemathesis/Pact-style checks and automated contract validation).
- Build reliable data-ingestion pipelines with robust retry/backoff and advisory-lock coordination for distributed workers, expanding the Celery + advisory-lock patterns already in the price-ingest service.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories: