Turn vague AI ambition into shipped product.
Peak3 specialises in 0-1 data-heavy & AI products and features. We help B2B & SaaS teams work out what to build, what to avoid, and how to ship quickly and securely.
Led AI product work across fintech, HR-tech, and carbon markets.
Example: AI product scope & delivery
- 01
Map user workflows
Identify where AI can remove friction or create leverage.
- 02
Pick the wedge
Prioritise by value, feasibility, risk, and data readiness.
- 03
Design the system
UX, data flows, evals, guardrails, fallbacks, launch plan.
- 04
Ship the product
We can ship the product with Claude Code securely and correctly, or work alongside engineering.
Experience across AI, B2B SaaS, fintech, HR-tech, and carbon markets
The problem
Most AI roadmaps fail before engineering starts.
Today, anyone can build with AI. The skill is in understanding the user, finding the right wedge, designing the product experience, and knowing how to evaluate whether the system is actually useful.
Too many vague ideas
Teams know they should be using AI, but the roadmap turns into a list of generic copilots, chatbots, and automation ideas with no clear commercial priority.
Too many prototypes
AI demos are easy. Useful AI products are harder. Most prototypes break when they meet real customer workflows, messy data, edge cases, and trust requirements.
Not enough product ownership
Engineering can build the system, but someone still needs to define the use case, user experience, success metrics, evaluation approach, rollout plan, and trade-offs.
What Peak3 does
Peak3 sits between strategy, product, and build.
We work with founders, product leaders, and engineering teams to move from unclear AI opportunities to focused product decisions, MVPs, V1 AI features, and shipped platforms.
AI opportunity mapping
Find the workflows where AI could create measurable value, not just novelty.
Use case prioritisation
Rank AI ideas by customer pain, revenue impact, data readiness, feasibility, and risk.
Product strategy
Turn scattered AI ideas into a clear roadmap, wedge, and sequencing plan.
AI feature design
Define the user experience, system behaviour, states, fallbacks, and human-in-the-loop moments.
Build leadership
We build with Claude Code or work with engineering to get it shipped without losing product clarity.
Evals and guardrails
Set up lightweight evaluation, quality checks, risk controls, and feedback loops.
Offers
Three ways to work together
Start with a focused strategy sprint, move into an MVP build sprint, or bring in senior AI product leadership for a few days a week.
AI Feasibility & Roadmap
A focused sprint to identify the highest-value AI opportunities in your product, assess feasibility, and leave with a practical 90-day roadmap.
- Product and workflow audit
- AI opportunity map
- Use case prioritisation
- Data and feasibility review
- Risk and guardrail notes
- 90-day roadmap
- Build vs buy recommendation
Outcome
A clear decision on what to build first, what to avoid, and how to sequence the work.
AI Build Sprint
Turn one AI use case into a clear product spec, working prototype, or first shipped workflow with your engineering team.
- Product spec
- UX and workflow design
- MVP or full product build
- Engineering collaboration
- Evals and quality checks
- Launch plan
- Customer feedback loop
Outcome
A working AI feature, prototype, or MVP that can be tested with real users.
Senior AI Product Lead
Senior AI product leadership for teams that need product direction, discovery, specs, prioritisation, and delivery support without hiring full-time.
- AI product roadmap
- Discovery and customer calls
- Product specs and PRDs
- Engineering prioritisation
- Stakeholder management
- Launch and iteration
- Team coaching
Outcome
Senior product ownership across AI strategy, delivery, and commercial trade-offs.
Indicative pricing. Scope depends on complexity, delivery model, and whether technical build support is required.
Process
A practical process for shipping AI features
- 01
Diagnose
We map the customer workflow, business goal, data reality, and existing product constraints.
Outputs
- Workflow map
- User pain points
- Data sources
- Product constraints
- 02
Prioritise
We choose the use case that has the best mix of customer value, feasibility, and commercial impact.
Outputs
- Prioritised use case list
- Feasibility score
- Risk notes
- Recommended wedge
- 03
Design
We define how the product, AI feature or workflow behaves, including UX, aha moments, states and fallbacks.
Outputs
- Product spec
- User flows
- System behaviour
- Guardrail notes
- 04
Build
We ship quickly and securely with Claude Code, or work with engineering directly. We can also advise your development team on how to build quickly with AI.
Outputs
- Working product
- Technical handover
- Launch checklist
- Feedback loop
- 05
Evaluate
We test quality, usefulness, adoption, and failure modes before expanding the feature.
Outputs
- Evals
- Success metrics
- User feedback
- Next roadmap decision
Selected work
Case studies
Selected examples across AI product strategy, B2B SaaS, automation workflows, and data-heavy products.
Financial Intelligence · B2B SaaS · 0-to-1
9fin's ESG platform from zero to scale
Designed, built and commercialised a new product line at 9fin (now a unicorn) combining data ingestion and analytics with LLM document analysis. Scaled it into a category-defining product used across the top tier of global capital markets.
Relevant capabilities
- 0-to-1 product strategy
- AI-powered document analysis
- Vector search
- Enterprise customer discovery
- Commercial prioritisation
- GTM and sales support
- Hiring and team leadership
Outcome
9 of top 10 global investment banks adopted.
HR-tech · Series A · B2B SaaS
Spotted Zebra's first AI voice agent & AI-augmented workflow.
Product strategy and delivery for a new AI interview agent inside Spotted Zebra's hiring platform, from discovery through production. Co-designed an AI-Augmented Delivery model with the engineering lead: modern tooling and squad workflow redesign.
Relevant capabilities
- AI agent product strategy
- Conversational and voice design
- AI-Augmented Delivery methodology
- MCP and modern tooling workflows
- Product requirements and scoping
- Engineering and squad collaboration
Outcome
~80% reduction in development cycle time.
Climate · Carbon Markets · Sales Intelligence
Ca2's buyer-intelligence platform for carbon markets
Designed and built Ca2, a sales intelligence product for carbon project developers, combining registry data, entity resolution, structured buyer profiles and LLM matching logic to surface likely corporate buyers in a fragmented market.
Relevant capabilities
- Data ingestion and cleaning
- Carbon registry and market data
- Entity resolution
- Buyer profiling and matching
- Sales intelligence design
- 0-to-1 product strategy
Outcome
11k+ buyer records resolved and matched.
Why Peak3
Led by a Series A-C product leader, 0-1 entrepreneur and hands-on builder
Peak3 is led by Jack David, an AI-native product leader with 10 years of experience across 0-1 entrepreneurship, B2B SaaS and Big 4 consulting including at one of Europe's fastest-growing unicorns.
- Led the ESG AI platform from 0 to scale at 9fin
- Worked across AI product and B2B SaaS delivery
- Built AI products and prototypes using LLMs, RAG, data pipelines, evals, and automation workflows
- Works with existing engineering, design, and commercial teams
- Strong at turning messy ideas into product direction, specs, and shipped workflows
- Comfortable with commercial trade-offs, customer discovery, pricing, and GTM
10 yrs
Product, consulting, and B2B SaaS
$1m ARR
0 to 1 product build
AI-native
LLM workflows, evals, RAG, automation
B2B SaaS
Fintech, HR-tech, carbon markets
Best fit
Who Peak3 works best with
Peak3 is best suited to B2B SaaS teams that have real product complexity, real customers, and a serious reason to explore AI.
Good fit
- Series A to C B2B SaaS companies
- Founder-led teams exploring AI product opportunities
- Product teams with too many AI ideas and no clear sequence
- Companies with workflow, data, or document-heavy products
- Teams that need senior product leadership but not a full-time hire
- Businesses that want to ship something useful within 6–10 weeks
Not a fit
- Internal AI training, workshops, or ChatGPT rollouts
- Consumer, D2C, or hardware-led products
- Pure ML research, model training, or internal infra projects
- Teams expecting an outsourced engineering team rather than embedded product leadership
Are you an agency?
No. Peak3 is a specialist AI product consulting and studio practice. I usually work directly with founders, product leaders, and engineering teams as a senior product lead, not as a traditional agency layer.Do you build the product or only advise?
Both, depending on the engagement. I lead product strategy, discovery, specs, workflows, and delivery. For build sprints, Peak3 can build the MVP directly or work with your engineering team where that is the better route.Are you a developer?
No. I am not positioning as a full-stack engineer. I am technical enough to work closely with engineers, prototype with modern AI tools, and make good product and architecture trade-offs, but my core value is product leadership.What types of AI products do you work on?
Workflow automation, internal copilots, customer-facing AI features, data products, document intelligence, voice agents, RAG systems, LLM-powered analysis tools, and AI-assisted operational workflows.What stage companies are the best fit?
Usually Seed to Series C B2B SaaS companies with existing customers, a product team, and enough technical capacity to ship. I can also support founder-led teams before they make their first senior product hire.Can you work with our existing engineering team?
Yes. That is usually the best setup. I help define the product direction, user experience, trade-offs, and delivery priorities so engineering can move faster with less ambiguity.Do you work internationally?
Yes. Peak3 works remotely with UK, European, and US teams. CET-friendly hours are preferred.Can you sign an NDA?
Yes. Most work can be covered by a standard NDA and services agreement. I also use IP carve-outs for reusable frameworks, playbooks, and pre-existing tools.
Have an AI product idea but no clear path to shipping it?
Book a short call to discuss where AI could create value in your product, what is worth building first, and whether Peak3 is the right fit.