LLM workflow systems
Prompt-driven pipelines, AI-assisted QA, source-backed validation, quarantine flows, rollback scripts, content review queues, and safety gates.
AI Automation Engineer
Daniel Pogodin builds practical AI systems for real operations.
I build LLM workflows, Python automation, source-backed validation, review queues, full-stack platforms, and tools that reduce operational friction without losing human control.
5+
years leading technical projects and delivery.
1000+
question EdTech systems with audits and premium access.
140
local pytest checks on the Job Agent automation system.
2026
active AI, automation, research, and data tooling.
What I build
Automation that is useful because it is constrained, testable, and reviewable.
My work is centered on turning messy operational workflows into structured systems: ingest data, normalize it, score it, generate outputs, route uncertainty to humans, and prevent unsafe or fabricated actions.
Python automation
Data imports, scoring, scraping support, Playwright-oriented scanning plans, Gmail draft generation, duplicate prevention, and guarded execution.
Full-stack products
React, TypeScript, Supabase/PostgreSQL, auth, payments, dashboards, progress tracking, entitlement checks, deployment, and smoke testing.
Selected projects
AI-assisted products, data tooling, and technical education platforms.
2026
AI automation
Job Agent
AI-assisted job-search and application automation system built around reviewability and submission safety.
- Imports job leads, normalizes records, scores fit, generates outreach drafts, and exports review queues.
- Includes ATS detection, form scanning plans, do-not-contact rules, duplicate prevention, and submit guards.
- Guarded Gmail draft creation with anti-fabrication controls and local test coverage.
PythonPlaywrightBeautifulSouppandasSQLiteGmail APIpytest
2025-2026
EdTech / LLM content
Amiram Academy / AmirNet
AI-powered English exam preparation platform with practice flows, premium access, and content operations.
- Built React/TypeScript/Supabase product flows for practice, progress, coupons, payments, and dashboards.
- Created AI-assisted pipelines for vocabulary, grammar, reading, mixed review, and listening workflows.
- Added validation, audit, import/export, quarantine, rollback, and smoke scripts.
ReactTypeScriptSupabasePostgreSQLClaude APIOpenAI APIPlaywright
2025-2026
Technical education
TDelectric / TD Academy
Electrical licensing exam simulator and learning product with source-backed quality controls.
- Delivered 1000+ questions, timed simulations, progress tracking, authentication, premium access, and cloud sync.
- Implemented official electricity-law imports, audits, review workflows, committee data, and release gates.
- Added CardCom entitlement checks, premium-security scripts, Sentry, Vitest, and uvu coverage.
ReactViteSupabaseFirebaseGoogle GenAICardComSentry
2026
Data research
Market Intelligence Labs
Research systems for wallet intelligence, trading strategy monitoring, and controlled paper execution.
- Built Polymarket ingestion, wallet tracking, trader ranking, edge detection, watchlists, and signal generation.
- Built SPY Sweep MSS infrastructure for data ingestion, feature generation, replay validation, OOS checks, and reports.
- Used risk gates, paper/shadow monitoring, regression tests, and exchange-calendar-aware workflows.
TypeScriptNode.jsPostgreSQLClickHousePythonDuckDBPlotly
Good automation depends on clear inputs, validation, review gates, tests, and handover quality. That is the discipline I bring to AI-assisted products.
Technical stack
A practical stack for building, testing, and operating AI-assisted products.
AI and automation
LLM workflowsClaudeOpenAIGoogle GenAIAI QAHuman reviewSafety gates
Languages
PythonTypeScriptJavaScriptSQLHTMLCSS
Product
ReactViteTailwindshadcn/uiSupabaseFirebaseNetlifyPayments
Data and quality
pandasDuckDBSQLiteClickHousePostgreSQLPlaywrightpytestVitest
Delivery background
Before AI automation, I delivered technical projects with real budgets, schedules, teams, and field constraints.
That operating background is why my software work emphasizes reliability, operational fit, and clear approval paths instead of demos that only work in ideal conditions.
Project Manager
Managed a strategic electrical project end to end, including planning, execution, control, handover, workforce coordination, supplier/client negotiations, contract analysis, pricing, and AI-assisted tooling for pricing and process quality.
Project Manager
Managed renewable-energy projects from planning through delivery, with responsibility for profitability, quality, schedule, budget, customer coordination, specifications, planning review, and execution controls.
Electrical and Controls Engineer
Designed electrical and control panels for large projects with Siemens, Allen-Bradley, and Sauter environments, handled PLC-related tasks, and ran on-site troubleshooting.
Electrical Engineer and Panel Draftsman
Designed electrical panels and schematics, and coordinated with the Standards Institute, customers, and suppliers.
Contact
Useful for teams that need automation shipped with judgment, tests, and operational controls.
Open to AI automation, LLM application, Python automation, and full-stack automation roles. Best fit: teams building practical internal systems, workflow automation, data validation, EdTech, or operational tools around AI.