01Working prototype
Every story starts with proof.
Approved facts, traceable claims, and owned media enter as one source package, so the creative work begins from something the team can inspect.
← Back to studio
Engineering
Systems, code, and honest labels.
The technical door — proof, systems, experience, and skills with status labels kept in plain sight.
Current flagship · private dry-run
Build in progressOne approved brief. A reviewable video campaign.
Content Engine is an evidence-first video operations system for teams that need more content without handing brand, budget, or publishing to a black box. It connects planning, production, quality checks, and partner review in one visible workflow.
- Working prototype
- Human approval
- Responsive workspace
- Private preview
At a glance
How one brief moves through Content Engine.
A plain-language view of the same six gated stages shown in the full case-study flow.
- 01Brief + sourceApproved facts and owned media enter together.
- 02StrategyAngles and scripts stay tied to the brief.
- 03ProductionLong- and short-form assets move as one package.
- 04QC + disclosureDeterministic checks create visible evidence.
- 05Human reviewPartners request, watch, and comment in context.
- 06Controlled releaseThe final decision stays explicit and human.
My role · Product architecture · backend orchestration · partner UX · release guardrails · creative workflow
30-second product film
See one brief become a review-ready campaign.
A concise walkthrough of how source material, production, multiple formats, contextual feedback, and human release controls stay connected.
Read the English transcriptThe film includes burned-in captions. An optional English caption track and written transcript are also available.
One approved brief enters Content Engine.
The dashboard keeps every source, script, and decision connected.
From that single story, it prepares one long-form film
and five vertical cuts, held together as one review package.
Partners watch, compare, and leave feedback at the exact moment.
Quality, rights, AI disclosure, cost, and approval stay visible. Automation moves the work. A person decides when it leaves.
The visible loop
Four chapters · one controlled campaign02Verified internal render
One brief becomes a campaign.
A verified internal run produced one 101-second film and five 34–41-second vertical cuts, kept together as one review package.
03Staged partner preview
Review happens in context.
Partners can request work, watch versions, see status, and leave timestamped feedback from one calm workspace on desktop or mobile.
04Human-gated release
The machine moves. People decide.
Quality, disclosure, rights, cost, and approval remain visible. Missing required evidence keeps external release closed.
Working private prototype · internal render and QA evidence shown · no public release or customer-performance result is claimed.
Full case study
Want the complete system view?
Open the business value, six-stage workflow, verified preview output, controls, stack, and release-specific QA in a focused reading window.Private prototype · no public-performance claim
02
Systems
Voice receptionist case study, archived system blueprints, and a live pipeline snapshot.
AI Voice Receptionist — Edwards Pipe & Tobacco
A voice agent answering over a 2,600-product catalog via RAG — Twilio telephony, ElevenLabs voice, and retrieval over the product database.
What happens on a call
A caller asks. The AI finds the answer.
The receptionist handles the conversation and checks the catalog before it answers.
- 01Phone ringsA customer calls the store.
- 02AI answersThe voice agent understands the request.
- 03Catalog searchedIt checks 2,600+ product records.
- 04Call resolvedThe caller gets a useful answer.~80% resolved
Under the hood
Live call · illustrative
caller · Do you carry Peterson pipes?
agent · Checks the 2,600-product catalog → answers → call resolved.
TwilioElevenLabsRAG over product DB
Honest framing: case-study metrics are presented as reported, not as mass-rollout benchmarks.
~80%
of calls resolved by the agent
71%
positive caller sentiment
2,600+
products in the RAG catalog
B2B Prospecting Pipeline
LangGraph agent pipeline, self-hosted on a VPS with systemd timers and a worker service. No public demo — it runs outreach — so the architecture stands in.
KvK / SBI filter
Hunter.io enrich
Apify scrape
LangGraph orchestrator · worker service
Automated outreach
VPS · systemd timers · cron
server-1 · prospect-worker
$ systemctl is-active tars-prospector.timeractive$ curl -s localhost:8088/health{"ok":true,"jobs":["prospect_research"]}$ journalctl -u tars-prospector -n 2REFILL added=2 backlog_total=98CYCLE done dispatched=0 backlog_total=96
One operating system
I work where story, systems, and delivery meet.
The creative work and the engineering work solve the same problem from opposite ends. One earns attention; the other captures, routes, and compounds it. I direct the story, build the system underneath it, and connect both to the operation that has to deliver.
That means I can translate between the people shaping the idea, the people using the workflow, and the technical system underneath it — without losing sight of the business decision the work is meant to support.
I can take a story from storyboard to system map and keep the same business decision visible the whole way through.
Experience
Where the range came from.
Jan 2026 — Present
Founder — AI Systems & Creative Operations — 180 Meridian
AI systems & creative studio · Rotterdam / Remote
Lead Content Engine, an evidence-first video operations platform that joins structured inputs, long- and short-form production, responsive partner review, and human-gated release; also build voice, retrieval, prospecting, and commerce systems.
Apr 2024 — Present
Creative Director / Business Operations Lead — All Around Universal
Remote · Denver market
Lead marketing, e-commerce, and automation initiatives with a team of 5 across websites, social campaigns, ads, and workflows; [OWNER: confirm] approximately €37K revenue managed in 2025.
Feb 2025 — Dec 2025
Lead Videographer & Marketing Strategist — The Little Local Book
Remote · Denver market
Produced short-form video, photography, and social assets for local businesses across [OWNER: confirm] hundreds of campaigns; planned content calendars and translated brand goals into creative execution.
2018 — 2024
Guest-Facing Hospitality — Monarch Casino & Resort · The Broadmoor
Colorado
Five-star service standards, fine-dining and high-volume.
Creative & marketing
Content operationsCampaign strategyAI video productionCreative directionShort-form socialBrand positioningfal.aiKlingVeoHeyGenElevenLabsNano Banana Pro
AI & LLM
LangGraphMulti-agent systemsMCP serversRAGpgvectorTool-callingStructured outputsPrompt engineeringEvalsClaudeGPTLlamaQwenDeepSeekOllama & vLLM self-hostingElevenLabsHeyGenTwiliofal.aiKlingNano Banana Pro
Engineering
PythonTypeScript / JavaScriptSQLFastAPINode.jsNext.jsREST / webhooksBullMQ / Redis queuesPostgreSQLSupabaseDockerCoolifySelf-hosted VPSHuman-in-the-loop systemsEvidence & release controlsPlaywrightpytestLinuxtmuxcron / systemdCIGoHighLevelStripeShopifyHunter.ioApify
Operations
CRM & pipeline managementSales & business developmentE-commerce & ShopifyGoHighLevelTeam leadershipProduction planningCampaign delivery
Education
B.A., Tourism & Hospitality Management — Turiba University, Riga, Latvia · 2016–2021
Japanese Philology — Tashkent Oriental Lyceum · 2013–2016
