All mentors

Raluca Bejan

GenAI Advisor: Zero to Production | Engineering Leader @ Adobe | Mentoring Startups on Scaling Teams & AI Systems

Bucharest, Romania (+02:00 UTC) Englishfrom Bucharest, Romania
Usually responds in 10 hours
Free
Price per hour
15 min30 min60 min
Time Blocks Available
5.00
7 reviews / 8 sessions
Fri
16
Next availability

Bio

Hi, I’m Raluca. I help startup teams ramp up in GenAI and turn prototypes into production-ready systems. New to GenAI? I'll help you pick viable use cases, avoid traps, and design an approach that fits. Already building? I can help with architecture, evaluation, safety, and monitoring. I bring both hands-on architecture depth and leadership experience. I’ve founded and scaled engineering teams, and I’m happy to coach engineers and managers while we ship. I can help you with: Use-case selection & product shape Identifying what’s viable, what’s risky, and what’s not worth building based on production realities. Agent architecture Designing state/memory systems, tool orchestration, and multi-agent workflows. MCP integration Implementing standardized tool/data access through MCP servers with a focus on reliability patterns. RAG done right Optimizing retrieval, grounding, and SOP execution through evaluation-driven iteration. Quality & evaluation Building infrastructure for golden datasets, automated scoring (LLM-as-judge), and tracing-driven debugging. Safety & governance Implementing guardrails, policy enforcement, and human-in-the-loop approvals for high-risk actions. Team building & execution Navigating hiring, scaling organizations, mentoring leaders, and resolving cross-functional friction. Just book a free call with me and let's chat, you've got nothing to lose.

Expertise

  • Artificial intelligence

    I’ve built production text-first agent systems that hold up under real constraints—agents vs workflows vs RAG, state/memory, tool calling, and orchestration (single/multi-agent; A2A/LangGraph). I also cover the NLP/LLM fundamentals behind reliability: prompting, grounding, routing, and context management.

  • Building a team

    I’ve scaled engineering teams and led the full people cycle: hiring, performance reviews, growth plans, and leadership coaching. I help founders build the operating cadence and ownership model needed to ship consistently under pressure, while navigating the cross-functional friction that builds up as teams grow.

  • Customer success

    I’ve built evaluation and observability loops that keep agent behavior stable for customers: regression/golden datasets, automated scoring, and tracing-driven debugging. This reduces customer-facing regressions, speeds up incident triage, and creates a continuous improvement loop so releases get better without breaking existing workflows.

  • Product launches

    I’ve shipped production systems and helped teams get from prototype to launch-ready: defining milestones and release gates, reducing delivery risks early, and putting reliable deployment in place: continuous delivery pipelines, environment management, and monitoring, so releases are repeatable, safe, and fast.

  • Product market fit

    I’ve built feasibility and use case frameworks that help teams find and validate the right GenAI product direction: where AI creates real user value, how to define success, and which constraints (cost, latency, risk) shape the solution, so you avoid expensive dead-ends and reach a testable first release fast.

  • Technology and tools

    I’ve built information service integrations that connect agents to real systems, APIs, databases, internal services, and SaaS tools, using clean contracts (i.e. MCP style), permissions, and failure handling, to ensure predictable execution.

Experience

  • Adobe

    Senior Machine Learning Engineer
    •adobe.com

    • Translate open-ended questions about agent quality, reliability, and production behavior into measurable success criteria, evaluation strategies, technical workstreams, architecture decisions, and delivery milestones. • Design and deliver distributed evaluation and developer-feedback workflows that connect agent traces, automated evaluations, human feedback, and production failures with scoring, regression testing, quality gates, and prioritized engineering improvements. • Built production diagnostic agents that integrate APIs, databases, notebooks, observability systems, and enterprise tooling to gather evidence, generate root-cause hypotheses, recommend actions, and support operational workflows from problem definition through production adoption. • Establish evaluation-driven engineering practices for assessing agent behavior, identifying failure modes, measuring production readiness, and converting findings into actionable improvements for engineering teams. Key areas: Agentic AI, AI evaluation, LLM reliability, developer platforms, MCP, A2A, LangGraph, OpenAI SDK, RAG, distributed systems, observability, and enterprise AI.

  • Adobe

    Sr. Engineering Manager
    •adobe.com

    Sr. Engineering Manager Stitching in Adobe Analytics • Owned technical strategy, roadmap, architecture, and engineering execution for distributed identity and analytics systems processing billions of events per day at petabyte scale. • Grew and led an engineering organization of more than 20 people, establishing scalable operating procedures, setting priorities, and coordinating complex dependencies across distributed engineering, product, platform, privacy, and data-quality teams. • Defined and scaled engineering processes for onboarding, planning, delivery, production readiness, release governance, on-call operations, incident response, and continuous improvement, enabling new teams and products to operate reliably at enterprise scale. • Established operational-readiness standards for new production capabilities, including service ownership, observability, support models, reliability requirements, DevOps practices, and service-level objectives and agreements. • Partnered with senior product and engineering leadership on technical strategy, organizational planning, delivery risk, platform evolution, reliability, privacy, and scale. • Helped evolve Cross-Device Analytics into capabilities supporting Adobe Customer Journey Analytics for global enterprise customers. • Built hiring, interview, onboarding, coaching, performance-management, and leadership-development practices. Designed an interview framework adopted across five teams and conducted more than 100 technical interviews.

  • Adobe

    Engineering Manager
    •adobe.com

    Engineering Manager Cross-Device Analytics • Founded and scaled a new engineering organization from 3 to 20 engineers in a fast-moving, 0-to-1 product environment. • Drove engineering execution for Adobe’s first cross-device identity and attribution capabilities, translating architectural direction into delivery plans, team priorities, and production outcomes. • Translated ambiguous enterprise and product requirements into technical strategy, milestones, staffing plans, release plans, and scalable platform capabilities. • Established engineering, hiring, delivery, mentoring, and operational practices for the new organization.

  • Luxoft

    Project Manager/Scrum Master

    Driving operational & administrative projects within UBS Global Equity Derivatives division • Coordinated delivery across more than 30 engineers working on mission-critical capital-markets platforms. • Partnered with front-office, risk, engineering, and architecture stakeholders to translate complex business and technical requirements into delivery plans, milestones, dependencies, and release execution. • Managed cross-team risks, delivery visibility, stakeholder alignment, and execution within a highly regulated environment.

  • Capgemini

    Scrum Master

    Led Agile delivery for enterprise sourcing and procurement platforms across distributed engineering and business teams- http://www.capgemini.com/services-and-solutions/outsourcing/business-process-outsourcing/solutions/ibx-procurement-cloud/. • Coordinated planning, execution, dependencies, delivery risks, and stakeholder communication for teams of more than 20 people. • Improved delivery visibility, cross-functional collaboration, issue resolution, and execution cadence. • Partnered with technical and business stakeholders to translate product objectives into clear priorities and incremental delivery plans.

  • BitDefender

    Technical Project Manager

    Coordinating the development of the product to meet the objectives, applied metrics, created project plans, identified, documented, and managed scope and deliverable change requests required whether initiated internally or by the client

  • BitDefender

    Technical Lead

    Participated in development of architecture and formulate detailed specifications and tasks. Worked directly with the manager from the testing team (working both sides on procedures for solving issues and bugs and taking corrective and preventive actions as well as developing validations rules) as well as with support and documentation teams. Taking part in planning, executing (manage project execution and information distribution) and monitoring and controlling (monitor and control project work, integrated change control, schedule control, manage project team).

  • BitDefender

    Software Programmer

    Understood functional requirements and participated in the architectural decisions, developed new features and assured maintenance for the existing features.

My best written content

Made within Glyfada