This is a fully remote role, open to candidates in North America, LATAM, Europe, Asia and the Middle East.
Company: Stealth-Mode Marketplace Startup | MENA Region
We're a live commerce and social marketplace built for the Middle East. We combine livestream shopping, social engagement, real-time auctions, direct product listings, seller tools, secure checkout, and buyer protection into one marketplace experience.
We are building a highly automated, data-driven, and AI-powered platform where buyers receive personalized shopping experiences, sellers receive intelligent growth tools, and internal teams operate more efficiently through automation.
Our next major phase is to build the AI and data foundation that powers personalization, buyer and seller segmentation, lifecycle automation, marketing automation, lookalike campaigns, recommendations, seller intelligence, and operational automation.
About the Role We are looking for a highly hands-on Lead AI & Data Platform Engineer to own our site's AI, data platform, and growth automation infrastructure.
This is not a pure research role. We need someone who can design, build, deploy, measure, and improve production systems. You will work across data engineering, event tracking, AI automation, LLM integrations, recommendation systems, marketing data activation, lifecycle automation, and internal AI tools.
You will be responsible for turning raw marketplace activity into clean, structured, actionable intelligence that powers product decisions, buyer personalization, seller growth, automated marketing campaigns, lookalike audiences, CRM automation, notifications, and executive reporting.
You should be comfortable moving between architecture and implementation, choosing when to build internally, when to use open-source tools, and when third-party APIs make more business sense.
Key Responsibilities1. Data Platform & Central Warehouse- Architect, build, and manage our site's central data warehouse using ClickHouse or similar high-performance OLAP databases.
- Design scalable data models for buyers, sellers, livestreams, auctions, products, orders, payments, shipping, marketing attribution, notifications, and platform engagement.
- Build reliable pipelines that transform raw events into clean datasets, dashboards, segments, alerts, and automated workflows.
- Ensure the data warehouse becomes the single source of truth for product, growth, marketing, finance, seller success, and management reporting.
- Define data quality rules, validation checks, monitoring, and alerting for broken or missing event flows.
- Build clear data documentation so product, engineering, marketing, and leadership teams can understand and trust the data.
2. Event Tracking, Telemetry & Behavioral Data- Design and implement robust event tracking across web, iOS, Android, livestreams, auctions, checkout, seller tools, search, chat, notifications, and product interactions.
- Define event schemas, naming conventions, user identity resolution, session tracking, and cross-device behavior mapping.
- Build buyer and seller behavioral datasets from activity such as watch time, bids, purchases, follows, bookmarks, saved shows, viewed products, chat activity, category interest, seller interaction, and retention behavior.
- Work with engineering teams to ensure tracking is accurate, scalable, and privacy-aware.
- Build the foundation for advanced analytics, recommendation systems, personalization, lifecycle triggers, and growth automation.
3. Growth Data Activation & Paid Marketing Automation- Build the data infrastructure needed to activate high-quality buyer and seller segments across advertising, CRM, lifecycle marketing, and notification channels.
- Design automated audience pipelines from the central data warehouse into platforms such as Meta, Google, TikTok, Snapchat, email, push notification, SMS, WhatsApp, and CRM tools.
- Create buyer and seller segmentation models based on GMV, engagement, category interest, livestream activity, bidding behavior, purchase frequency, retention, seller quality, and trust signals.
- Build lookalike audience workflows using high-value buyers, repeat purchasers, category-specific buyers, livestream viewers, abandoned checkout users, VIP buyers, high-performing sellers, and retained users.
- Build attribution and feedback loops that connect campaign performance back into the data warehouse, allowing us to understand which channels, audiences, creatives, and campaigns drive real GMV, not just installs.
- Help marketing teams improve ROI by targeting better audiences, reducing wasted ad spend, personalizing campaigns, and identifying the highest-value cohorts.
- Support server-side tracking and conversion APIs for paid platforms where needed, including Meta CAPI, Google Enhanced Conversions, TikTok Events API, Snapchat CAPI, and offline conversion uploads.
4. Lifecycle Marketing Automation & In-App Personalization- Build behavior-based lifecycle automation across our platform using buyer, seller, product, category, livestream, bidding, and purchase data.
- Design trigger-based communication flows across push notifications, email, SMS, WhatsApp, and in-app messages.
- Create personalized recommendation triggers based on user behavior, including watched livestreams, followed sellers, saved shows, category interest, viewed products, bids placed, abandoned checkout, past purchases, and similar buyer behavior.
- Build timing intelligence to decide the best moment to send each message, such as before a relevant livestream starts, after a buyer shows intent, when a seller goes live, when a similar product is listed, or when a buyer is likely to return.
- Build recommendation logic for products, livestreams, sellers, categories, auctions, and offers.
- Create automated journeys for buyer activation, first purchase, second purchase, reactivation, VIP buyers, inactive buyers, category-based buyers, and high-intent livestream viewers.
- Create automated journeys for seller activation, first livestream, first sale, seller retention, seller quality improvement, and high-potential seller support.
- Build frequency capping, quiet hours, channel prioritization, message ranking, and suppression logic to avoid spamming users.
- Connect lifecycle campaigns back to the central data warehouse to measure open rates, click-through rates, conversion, GMV, repeat purchase, retention, unsubscribe behavior, and channel performance.
- Work with marketing and product teams to test which messages, channels, timings, and recommendations drive the highest conversion and retention.
- Build the data layer needed for AI-generated personalized content, such as dynamic product recommendations, livestream reminders, category alerts, seller updates, and personalized offers.
5. AI Engineering & LLM-Based Automation- Build production AI workflows that support seller onboarding, seller scoring, customer support routing, product listing improvement, content moderation assistance, campaign generation, and operational automation.
- Design and deploy LLM-based internal tools for support, seller success, marketing, product, and operations teams.
- Evaluate and integrate AI APIs, open-source models, vector databases, RAG workflows, agent frameworks, and model orchestration tools.
- Build AI systems with proper logging, evaluation, guardrails, fallback logic, human review workflows, and cost monitoring.
- Create reusable AI services and APIs that can be used across our platform.
- Keep AI features practical, measurable, and connected to business outcomes.
6. Personalization, Ranking & Recommendation Systems- Build recommendation and ranking logic for live shows, sellers, products, categories, search results, and notifications.
- Create personalization models based on buyer interests, behavior, purchase history, livestream watch time, bidding activity, followed sellers, category affinity, and similar users.
- Support For You style discovery experiences for live commerce.
- Build buyer and seller intelligence models that help us identify high-potential buyers, valuable sellers, churn risks, inactive users, and growth opportunities.
- Create scoring systems for buyer levels, seller levels, lifecycle stages, and trust-based segmentation.
- Work with product and growth teams to test and improve recommendation quality.
7. Live Commerce AI & Media Automation- Explore and build AI features for livestream workflows, including transcription, translation, summarization, content tagging, clip extraction, and moderation assistance.
- Work with real-time media systems such as LiveKit, WebRTC, audio/video pipelines, speech-to-text, and translation tools.
- Build automation that helps convert livestream content into reusable marketing assets, including short clips, product highlights, seller summaries, and campaign-ready content.
- Analyze livestream performance data to help sellers improve conversion, engagement, auction success, and viewer retention.
8. Programmatic SEO & Marketplace Content Intelligence- Support scalable SEO systems for marketplace listings, seller pages, product pages, livestream pages, category pages, and search landing pages.
- Use AI to improve multilingual product content, metadata, structured data, search relevance, and content quality.
- Build systems that identify high-opportunity categories, keywords, listings, and content gaps.
- Ensure AI-generated content is high-quality, brand-safe, multilingual, and aligned with platform standards.
9. MLOps, LLMOps & Production Reliability- Build the technical foundation for deploying, monitoring, evaluating, and improving AI systems in production.
- Implement prompt versioning, model evaluation, experiment tracking, cost monitoring, latency tracking, and output quality checks.
- Build observability around AI workflows, including errors, hallucination risk, user feedback, failed tasks, and fallback paths.
- Define when to use closed-source APIs, open-source models, fine-tuning, RAG, rule-based systems, or traditional ML.
- Ensure AI and data systems are scalable, secure, maintainable, and cost-efficient.
10. Data Governance, Privacy & Security- Implement role-based access control, data permissioning, sensitive data handling, and secure data workflows.
- Ensure marketing audiences, AI workflows, and user data pipelines follow consent, privacy, and governance best practices.
- Help define data retention, anonymization, audit logs, and access policies.
- Work with leadership to ensure data is useful without becoming risky, messy, or non-compliant.
11. Technical Leadership & Cross-Functional Ownership- Own the AI and data platform roadmap in partnership with product, engineering, marketing, seller success, and leadership.
- Translate business goals into technical systems and measurable outcomes.
- Make clear build-vs-buy recommendations for tools, models, infrastructure, and platforms.
- Mentor engineers and help create best practices for data, AI, tracking, personalization, and automation.
- Help us build a future AI & Data team as the company scales.
Must-Have Qualifications- 7+ years of professional experience in software engineering, data engineering, AI engineering, machine learning engineering, or data platform architecture.
- Strong Python experience.
- Strong SQL experience and ability to design clean, scalable data models.
- Hands-on experience building production data pipelines and analytics infrastructure.
- Experience with OLAP databases such as ClickHouse, BigQuery, Snowflake, Redshift, Apache Druid, or similar.
- Strong understanding of event tracking, telemetry architecture, user behavior data, identity resolution, and data quality.
- Experience integrating LLMs, AI APIs, or AI models into production systems.
- Experience building backend services, APIs, automation workflows, and data-driven systems.
- Strong understanding of data activation, segmentation, lifecycle marketing, attribution, and campaign measurement.
- Experience building behavior-based triggers, personalized notifications, or lifecycle automation.
- Ability to work with marketing and growth teams to turn data into better targeting, personalization, and ROI.
- Strong understanding of APIs, cloud infrastructure, containers, CI/CD, logging, monitoring, and production reliability.
- Ability to work independently in a fast-moving startup environment.
- Strong communication skills and ability to explain technical tradeoffs to founders, product, engineering, and marketing teams.
Preferred Qualifications- Experience with ClickHouse.
- Experience with marketplace, e-commerce, social commerce, livestreaming, auctions, consumer apps, or high-volume transactional platforms.
- Experience with recommendation systems, personalization, search ranking, buyer scoring, seller scoring, feed ranking, or notification ranking.
- Experience with marketing data activation, CDPs, reverse ETL, server-side tracking, and paid media audience pipelines.
- Experience sending warehouse-based audiences to Meta, Google, TikTok, Snapchat, CRM, email, push, SMS, or WhatsApp platforms.
- Experience with conversion APIs such as Meta CAPI, Google Enhanced Conversions, TikTok Events API, Snapchat CAPI, or offline conversions.
- Experience with attribution, cohort analysis, retention analysis, A/B testing, incrementality testing, ROAS, CAC, LTV, and funnel analysis.
- Experience building lifecycle marketing automation, behavioral triggers, and personalized notification systems.
- Experience with push notifications, email, SMS, WhatsApp, and in-app messaging workflows.
- Experience with frequency capping, quiet hours, send-time optimization, suppression logic, communication preferences, and channel prioritization.
- Experience with LiveKit, WebRTC, real-time audio/video processing, speech-to-text, translation pipelines, or livestream analytics.
- Experience with RAG, vector databases, AI agents, prompt engineering, model evaluation, and AI observability.
- Experience with Arabic/English multilingual systems, RTL products, or regional marketplace localization.
- Experience with data privacy, access control, consent management, and secure handling of user data.
Relevant Tools, Frameworks & Technologies Candidates do not need to know every tool below. Strong candidates should have hands-on experience with several of them or equivalent alternatives. Do not reject strong candidates only because they used a similar tool instead of the exact tool listed.
AreaRelevant Tools, Frameworks, or Similar AlternativesData Warehouse & DatabasesClickHouse, BigQuery, Snowflake, Redshift, PostgreSQL, MongoDB, Redis, Apache Druid
Data Pipelines & OrchestrationAirbyte, RudderStack, PostHog, Kafka, Redpanda, Spark, Flink, dbt, Dagster, Prefect, Airflow
CDP, Reverse ETL & Marketing ActivationRudderStack, Segment, Hightouch, Census, Customer.io, Braze, Iterable, HubSpot, Salesforce Marketing Cloud, Meta Ads API, Google Ads API, TikTok Ads API, Snapchat Marketing API
Lifecycle Messaging & NotificationsFirebase Cloud Messaging, OneSignal, Braze, Customer.io, Iterable, Airship, SendGrid, Twilio, WhatsApp Business API, SMS providers, email marketing platforms, in-app messaging systems
AI, LLM & Agent FrameworksOpenAI API, Anthropic API, Gemini API, Hugging Face, LangChain, LlamaIndex, CrewAI, AutoGen, Semantic Kernel, DSPy, vLLM, Ollama
Vector Databases & RAGpgvector, Qdrant, Weaviate, Milvus, Pinecone, Elasticsearch, OpenSearch
MLOps & LLMOpsMLflow, Weights & Biases, LangSmith, Promptfoo, OpenTelemetry, Evidently AI, Feast, model evaluation and monitoring tools
Backend & InfrastructurePython, FastAPI, Django, REST APIs, GraphQL, Docker, Kubernetes, Terraform, GitHub Actions, GitLab CI/CD, Linux, cloud infrastructure
Real-Time Media & LivestreamingLiveKit, WebRTC, FFmpeg, speech-to-text APIs, translation APIs, audio/video processing pipelines
What Success Looks LikeFirst 30 Days- Understand our product, data flows, AI goals, personalization needs, and marketing automation requirements.
- Audit existing event tracking, data quality, and infrastructure gaps.
- Define the AI and data platform roadmap.
- Identify the highest-priority buyer, seller, marketing, lifecycle, and operational automation use cases.
- Recommend the right build-vs-buy approach for data pipelines, CDP, reverse ETL, lifecycle messaging, and AI workflows.
First 60 Days- Improve or redesign key event tracking architecture.
- Build the first clean buyer and seller datasets in the central warehouse.
- Define initial buyer segments, seller segments, lifecycle stages, and behavior-based triggers.
- Set up the foundation for marketing audience activation and campaign feedback loops.
- Set up the foundation for push, email, SMS, WhatsApp, and in-app trigger-based campaigns.
- Ship at least one production AI or automation workflow.
First 90 Days- Launch the first automated audience segmentation system for buyers and sellers.
- Create high-value buyer and seller cohorts for lookalike campaigns.
- Build the first data flow from the warehouse into marketing, CRM, or lifecycle messaging platforms.
- Launch personalized notification flows for livestream reminders, followed sellers, category interest, abandoned checkout, recommended products, and reactivation.
- Create reporting that connects paid campaigns, push, email, SMS, WhatsApp, and in-app campaigns to conversion, GMV, retention, repeat purchase, and ROAS.
- Launch the foundation for personalization, recommendations, buyer leveling, or seller scoring.
- Implement frequency caps and suppression rules to prevent over-messaging users.
- Deliver clear documentation and dashboards that leadership, marketing, product, and seller success teams can actually use.
Who This Role Is For- Someone who enjoys building serious infrastructure and also understands business impact.
- Someone who can answer questions like: which buyer is likely to buy, which seller needs support, which product should be recommended, which notification should be sent now, and which campaign actually drove GMV.
- Someone who can turn messy marketplace data into working systems that improve growth, personalization, automation, and decision-making.
Never miss an opportunity! Create an alert based on the job you applied for.
