How We Secured Our First AI Projects
This presentation explores our journey into AI development. We'll share how we landed two significant projects despite having no portfolio.
We'll cover our approach with existing clients in public safety and e-commerce sectors.
DS
by Dima Semensky
Behind the Curtain: How This Presentation Was Born
The Walking Brainstorm
Captured my thoughts during a 30-minute, 2-mile walk with just an iPhone and headphones.
The Voice Recording Magic
Spoke ideas freely using the voice recorder app while strolling along.
The AI Transcription Act
Uploaded audio to Fireflies.ai which transformed spoken words into written text.
The Grand Finale
Exported to PDF, imported to Gamma, and voilà—presentation complete with minimal effort!
Starting from Zero
Research Phase
Invested one year of research with our CTO.
Internal Prototyping
Built several prototypes based on our own ideas.
Understanding Capabilities
Gained insights into current and potential AI capabilities.
Approaching Our First Client
1
Client Selection
Chose a SaaS company known for innovation.
2
Stealth Development
Built 2-3 ideas quickly without telling them.
3
Strategic Reveal
Showed prototypes during a regular meeting.
4
Problem Exploration
Proposed a deeper dive into their current challenges.
Jobs-to-be-Done Approach
Training Sessions
Spent two days training the client on JTBD methodology.
Opportunity Identification
Used JTBD to identify improvement opportunities with AI in mind.
Knowledge Sharing
Explained our understanding of AI capabilities.
Key lesson: You don't need to know everything. Just know a bit more than the client.
From Ideas to Development
Documentation
Generated 20 written ideas in a document.
Prioritisation
CEO selected top three problems to address.
Team Formation
Created three groups to drill down on each idea.
Deadline Setting
Set annual conference as the demo deadline.
Project Implementation
Standard Development
Used typical software development lifecycle. AI development is no different than custom development.
Management Demo
Presented to CEO and management team before conference.
Public Presentation
Presented from stage at client's conference despite smaller audience.
Positive Outcome
Received enough positive feedback to proceed with paid projects.
Three Successful AI Solutions
Interactive Help
Created a chatbot around their online help system.
Data Extraction
Built AI to extract information from scanned files quickly.
Data Conversation
Implemented a system for talking to their data.
We added more team members and launched one solution in production.
Second Client Success Story
Initial Resistance
CEO was not receptive due to AI's past reputation.
Compelling Demos
Showed several interesting demos to change perception.
Product Release
Developed and released a back-end product for catalog enrichment.
Key lesson: Initiative and willingness to invest your time puts you in a better position when needs arise.
Our AI Learning Journey
We've moved from small experiments to full company-wide adoption. The journey continues towards our ambitious vision.
Initial Investment
Late 2022: Our CEO with technical background led early explorations alongside CTO and senior management.
Limited Experimentation
A handful of developers tested capabilities. Began client conversations to identify opportunities.
Company-wide Push
January 2025: Mandated adoption of ChatGPT, Cursor, and Lovable across all departments.
Present Reality
Half our staff use AI tools daily. BAs, Marketing, and Pre-sales lead adoption.
Future Vision: Synteron
Full automation through our DevCom entity. A seamless integration of humans, AI agents, and data.