On this page

AI Product Manager, Agentic AI Consultant and Agentic AI Trainer

Hey, I’m Georgi. I’m a hands-on AI Product Manager.

I lead product strategy, discovery, and go-to-market work for new services and products, with or without AI.

My work connects product management, consulting, and training with a practical operating model. Explore the evidence through my products, portfolio, writing, talks, and learning resources.

Experience

Regions
  1. Europe
  2. Middle East
  3. United States
  4. Central Asia
Domain Experience
  • Generalist
  • Energy Analytics
  • Healthcare (Sleep Apnea)
  • Cybersecurity
  • Insurance
  • Engineering Analytics
  • Logistics
Training
  1. Internal and external SoftServe programs
  2. Partner test drives
  3. Enterprise security topics
  4. Independent Agentic AI training for business analysts and product managers

About the Creator

Georgi Naydenov

AI Product Manager at SoftServe, Agentic AI consultant and trainer, business analysis expert, mentor, and conference speaker. I work across product strategy, discovery, service definition, go-to-market, delivery, and measurement. My credentials include CBAP, PMP, TOGAF Enterprise Architecture Practitioner, and multiple Scrum.org certifications.

I support pre-sales work, develop AI-enabled services and products, and connect technology choices with clear business and customer value. I have worked with clients across Europe, the Middle East, the United States, and Central Asia, mainly in healthcare and also in energy analytics, cybersecurity, insurance, engineering analytics, and logistics. I am the creator of the Agentic AI Library.

Credentials

OrganisationCredentialAbbrev.
AnthropicClaude Certified Architect - ProfessionalValid · Aug 2027CCAR-P
The Open GroupEnterprise Architecture PractitionerTOGAF EA Practitioner
IIBACertified Business Analysis ProfessionalValid · Oct 2028CBAP
PMIProject Management ProfessionalValid · Apr 2028PMP

Built independently

Personal Projects and Products

Products I have planned and built independently to make AI knowledge, product architecture, and new technologies easier to explore and use.

agenticlibrary.net

Open-source knowledge product

Agentic AI Library

  • Curated library
  • Role-based paths
  • Frameworks and tools
  • Governance and research

A curated product for discovering agentic AI frameworks, tools, research, infrastructure, governance resources, and practical references by mission and role.

Open Agentic AI Library
kaggle.com

End-to-end vibe-coded application

Kaggle Agents Capstone Project

  • Knowledge base
  • Swagger
  • ReDoc
  • Karpathy Wiki
  • In-app harness

An end-to-end vibe-coded app with a knowledge base, Swagger and ReDoc API documentation, a Karpathy Wiki, and an in-app harness.

Open Kaggle write-up

How I Build AI Products

This is the operating model I use to identify opportunities and turn them into clearly defined services, structured projects, and viable products, with or without AI. It covers the full product lifecycle, from an initial idea, request, or pre-sales opportunity through discovery, service definition, product development, go-to-market, production monitoring, and the training and support needed for successful adoption.

  1. 01

    Opportunity and service definition

    I shape an early idea, request, or opportunity into a clear service or product direction. I have supported the full pre-sales process, from creating security, go-to-market, SOC, and NOC service decks to defining client value propositions, proposal scope, assumptions, boundaries, and early delivery estimates when information is limited.

    • Security, go-to-market, SOC, and NOC service decks
    • Client value proposition and opportunity story
    • Proposal scope, assumptions, and boundaries
    • Early work breakdown structure, effort estimate, and delivery inputs
  2. 02

    Discovery and solution fit

    I lead discovery to understand the business problem, users, current process, expected outcomes, and main constraints before selecting the product or technology approach.

    • Problem statement and desired outcomes
    • Stakeholder, user, and process analysis
    • Constraints, dependencies, and risks
    • Value, feasibility, and solution-fit assessment
  3. 03

    Product strategy and direction

    I set product direction by connecting customer and market insights with business goals, commercial value, and technology opportunities. I make clear choices about where to compete, which problems deserve investment, how the product should stand out, and which outcomes should guide the roadmap and go-to-market plan.

    • Market, customer, competitor, and technology insights
    • Product vision, positioning, value proposition, and business model
    • Strategic choices, investment priorities, risks, and trade-offs
    • Product objectives, roadmap themes, and go-to-market alignment
    • Strategic roadmap, go-to-market direction, and product-market-fit plan
  4. 04

    Delivery planning and implementation

    I keep product, design, engineering, and business teams aligned as the product moves from definition to delivery. I break the product direction into testable releases, support trade-off decisions, and keep delivery focused on user and business outcomes.

    • Minimum viable product and release scope
    • User journeys and story maps
    • User stories and acceptance criteria
    • BPMN, sequence, and ArchiMate diagrams
    • Process, data, and system models
  5. 05

    Go-to-Market, Measurement, and Growth

    I define how the product enters the market, proves its value, and moves towards product-market fit. After launch, I connect adoption, engagement, retention, commercial performance, product quality, and production reliability to the decisions that support product growth. I also have experience defining observability and production monitoring with Grafana, Prometheus, and Splunk.

    • Go-to-market readiness and product adoption plan
    • North-star metric, KPIs, baselines, and owners
    • Product-market-fit signals, user feedback, and learning cycles
    • Product analytics, production monitoring, dashboards, and alerts
    • Evidence for roadmap priorities, improvement, growth, and scaling
  6. 06

    Training and enablement

    I create and deliver practical training for teams, partners, business analysts, and product managers. This includes GenAI test drives, enterprise security topics, AI adoption, and Agentic AI practice.

    • Internal and external SoftServe programs
    • Partner test drives and demonstrations
    • Deployment trade-off guidance
    • Role-specific Agentic AI training
    15 to 350
    people per training cohort
    93%
    training NPS

Extension of discovery and solution fit

Choosing the Right Product Approach

Discovery continues by identifying which product and technology approach best fits the problem, users, working environment, and expected outcomes. The right solution may be a web or mobile product, a data platform, rule-based automation, an AI-enabled workflow, or a multi-agent system. I select AI only when it provides clear and measurable value compared with a simpler approach.

Web, mobile, or data platform

Users and teams need a clear digital experience, reliable data, and predictable application or platform behaviour.

Choose software or a data platform when the workflow and data model are known and intelligence is not the constraint.

Automation and workflows

The process is repetitive, rule-based, and spread across people or systems.

Choose automation when the steps can be expressed and exceptions stay manageable.

AI-enabled workflow

People need help with language, knowledge, recommendations, or context-heavy decisions.

Choose AI when evaluation and human oversight can keep the outcome useful and safe.

Multi-agent system

The work spans specialist roles, tools, and dependent decisions that must be coordinated.

Choose multiple agents only when one workflow or one agent cannot handle the operating model.

I assess cloud and on-premises deployment options based on security, data sovereignty, latency, integration, performance, scalability, and cost requirements.

Extension of go-to-market, measurement, and growth

Measuring Product Performance

Once the product approach has been selected, measurement shows whether the product is reaching its market, delivering value, and moving towards product-market fit. I connect acquisition, activation, engagement, retention, revenue, and referral metrics with delivery performance, product quality, and production reliability. For AI-enabled products, I also measure evaluation quality, task success, failure patterns, human escalation, operating cost, and system health. These measures provide evidence for go-to-market decisions, roadmap priorities, improvement, and scaling.

Acquisition9 metrics

Shows how people find the product, which channels bring the right users, and how efficiently the go-to-market plan creates demand.

  • Bounce rate
  • Conversion rate
  • Landing page conversion rate
  • Customer acquisition cost (CAC)
  • Cost per acquisition (CPA)
  • Channel effectiveness
  • Traffic source distribution
  • Cost per click (CPC)
  • Click-through rate (CTR)
Activation7 metrics

Shows whether new users reach their first useful outcome and where onboarding prevents them from seeing the product's value.

  • Time to value (TTV)
  • Onboarding completion rate
  • User activation rate
  • Trial-to-paid conversion rate
  • First-time user conversion rate
  • Product qualified leads (PQL)
  • Product qualified accounts (PQA)
Engagement10 metrics

Shows how often people use the product, which features support their work, and whether they can complete important tasks successfully.

  • Daily active users (DAU)
  • Monthly active users (MAU)
  • Stickiness (DAU / MAU)
  • User satisfaction (CSAT)
  • Session length
  • Session frequency
  • Feature usage
  • Customer effort score (CES)
  • Task success rate
  • User feedback score
Retention7 metrics

Shows whether the product continues to create value over time and helps identify the users or groups most likely to stay or leave.

  • Churn rate
  • User retention rate
  • User renewal rate
  • Customer lifetime
  • Customer health score
  • Product adoption rate
  • Cohort analysis
Revenue14 metrics

Connects product use with commercial results and shows whether growth is sustainable after acquisition, delivery, and operating costs.

  • Total revenue
  • Monthly recurring revenue (MRR)
  • Annual recurring revenue (ARR)
  • Average revenue per account (ARPA)
  • Customer lifetime value (CLV / LTV)
  • Customer profitability
  • Expansion revenue
  • Net revenue churn
  • Net revenue retention
  • Average contract value (ACV)
  • Gross margin
  • LTV to CAC ratio
  • CAC payback period
  • Return on investment (ROI)
Referral4 metrics

Shows whether users are willing to recommend the product and whether referrals create a reliable source of new customers.

  • Virality coefficient
  • Customer referral rate
  • Referral conversion rate
  • Net promoter score (NPS)
Lean and Agile delivery8 metrics

Shows how quickly the team can deliver, learn, and respond to evidence without creating delays or too much work in progress.

  • Lead time
  • Time to market (TTM)
  • Cycle time
  • Work in progress (WIP)
  • Throughput
  • Time to learn (TTL)
  • Work item age
  • Velocity
AI product quality12 metrics

Shows whether the AI completes its task accurately, safely, and at an acceptable cost for the product and its users.

  • App-specific failure-mode rate
  • Failure frequency
  • Failure impact
  • Specification failure rate
  • Generalization failure rate
  • Task success rate
  • Relevance
  • Groundedness
  • Hallucination rate
  • Human escalation rate
  • Evaluation pass rate
  • Cost per successful outcome
AI evaluator quality6 metrics

Shows whether automated evaluations agree with human judgement and can be trusted to support product quality decisions.

  • Human and evaluator agreement
  • True positive rate (TPR / recall)
  • True negative rate (TNR)
  • Precision
  • F1 score
  • Evaluation cost
Reliability and observability14 metrics

Shows whether the product is available, stable, and healthy in production, and whether teams can find and resolve problems quickly.

  • Availability and uptime
  • Latency
  • Request throughput
  • Error rate
  • Failure rate
  • Incident rate
  • Mean time to detect (MTTD)
  • Mean time to recover (MTTR)
  • Resource utilization
  • Saturation
  • Service level indicators (SLI)
  • Service level objectives (SLO)
  • Log, metric, and trace coverage
  • Alert volume and noise