Choose the right AI for your payments and finance operations

I advise payments companies, fintechs, merchants and platforms on AI adoption and economics

Book a callMartin KoderischMartin KoderischAI adoption and economics adviser

What I do

I've spent over 20 years in payments, at Citi, Mastercard and Edgar, Dunn & Company. Whether the client was a PSP, an EMI or a merchant, the bottlenecks frequently sat in the same back-office processes, such as chargebacks and settlement reconciliation.

Today, AI can take on much of the routine casework in these processes, and what works in one firm largely carries over to the next. But adoption is lagging behind these rapid advances in AI capabilities. Hence, I specialise in AI adoption for these processes, helping payments and finance operations teams across all types of firm, from picking the first process to getting it into production.

I also run ScalePoint Research to back this up. It tests AI models across 11 clusters of these processes, including what each case costs at volume.

Issuer or bank
Acquirer or PSP
Merchant
Platform or marketplace
EMI or cross-border
Stablecoin or crypto

Payments and finance operations

ChargebacksSettlement reconciliationMerchant onboardingException handlingSanctions alert reviewCard statement checks+ 44 more →
The same back-office processes run in every type of firm. I work across all six.

Services

  1. 1

    Find the use case

    AI business case

    Which process should we start with, and does the business case stack up?

    2 to 3 weeks
  2. 2

    Choose the route

    Build, buy or hybrid

    Which software, models and partner should we use, and will it work on our cases?

    3 to 4 week sprint
  3. 3

    Pick the tools

    Software and partner selection

    Which vendor or partner should we pick, and what goes in the RFP?

    Shortlist and RFP pack
  4. 4

    Check the value

    Value check

    Which pilots should go into production, and are live tools delivering what was promised?

    Scoped to your tools
Alongside every step

Fractional Head of AI adoption

A few days a month to run AI adoption alongside your team, from picking use cases to moving pilots into production.

Training and speaking

Workshops, training and talks

Sessions for teams and conferences on AI in payments and finance operations, such as how to test models on your own cases.

For vendors

Independent evaluation

ScalePoint Research tests your product against your own claims and writes it up in a report your sales team can share, labelled as commissioned.

How independence works →
For vendors

Research reports

Market analysis, customer research and surveys your marketing team can publish under your brand, with ScalePoint Research credited as the research partner.

Talk it through

I'm also happy to discuss what you need and scope a tailored project.

Backed by ScalePoint Research

S1, reading card statements: 25 usable models

0%1%2%3%0.1¢1¢10¢Price per statement (log scale) →↓ Fees wrong, plain promptClaude Opus 5.5Grok 4.7Gemini 3.8 FlashGLM 5.3Mistral Large 3GPT-5.5DeepSeek V4.1 Flash
0%1%2%3%0.1¢1¢10¢Price per statement (log scale) →↓ Fees wrong, plain promptClaude Opus 5.5Gemini 3.8 FlashMistral Large 3
Top left is cheap and accurate. Each dot is one model, plain prompt.

The projects above use the same tests and economic models as these studies.

Judgement calls in payments operationsS3 · Study → Reconciling settlementsS2 · Study → Reading card statementsS1 · Study →
Explore the research

Latest writing

All writing →
Note · AI economics · Oct 2026Who owns the cost of AI errors, IT or operations?Firms can now pick a model for each step without leaving their cloud provider, and the cost of AI errors lands in the operations budget. Article · Models · Oct 2026Test design: what the model does and what code doesWhich steps the tests give to a model, which they leave to code, and the judgement skills they score. Article · AI economics · Oct 2026The Unit Economics of AI in Payments and Finance OperationsWhat an AI model really costs per case once checking, mistakes and fixed costs are counted, with figures from S1 and S2.

Survey

How are firms approaching AI in payments and finance operations? ScalePoint Research is asking COOs, CFOs and heads of payments or finance operations. It takes about 7 minutes, and everyone who takes part gets the findings.

Take the survey

Background

Martin Koderisch

I spent ten years in operating roles at Citi, Mastercard and the start-up Luup, then nine years at Edgar, Dunn & Company, latterly as a Principal, on more than 75 engagements for banks, fintechs, card networks, private equity firms and merchants, from setting strategy to rolling out new products and processes. Since 2024 I've worked independently with payments and fintech businesses on strategic finance and value creation. I split my time between London and Barcelona.

I work project by project or as a fractional Head of AI adoption.

You can also connect on LinkedIn.

Start with a short call

After the call, you get a one-page proposal with the shape of the project and a ballpark price. There's no fee and no commitment.

Book a callor email martin@scalepointpartners.com

You can also connect on LinkedIn.