Agentic Squads · Accountable AI · Owned Autonomy

Deploy agentic talent now.
Own the autonomy that follows.

Accountable for every action an agent takes; sovereign over the stack it runs on — we call it Owned Autonomy. Start where you need to: deploy senior agentic engineers into your squads now, or transform the whole bank. Either way, the people who built the demonstration build with you.

Agents are crossing from advice to action.
The real question is whether you can answer for it.

Adopting AI tools is no longer the hard part. The hard part is acting through them safely: a regulated bank can only let an agent move money or run a journey if it is accountable for every action that agent takes, and sovereign over the agents, data and models it runs on. You cannot be accountable for what you do not control.

That is accountable banking, and it is what we build with you: deploy agentic product squads, ship real AI propositions, stand up the platforms they run on — and leave the capability with you. Not advice. Not a vendor's black-box OS you rent.

DTA Model: Teams and Capability

One ladder. Start anywhere.

Hire the talent, build a proposition, or transform the bank — each rung stands alone, and each leads to the next. Senior, control-grade, outcome-led.

Agentic Talent & Squads

1 · Agentic Talent & Squads

Deploy senior agentic engineers — embedded in your teams or as a managed pod — productive from week one. Not generic contractors: builders fluent in agentic frameworks, evals and on-device models, working to accountable, control-grade standards because they're led by people who've run bank transformation. The fastest way to start.

  • Senior agentic engineers — embedded or a managed squad
  • Productive from week one, inside your codebase and guardrails
  • Control-grade by habit — accountable delivery, not body-shop hours
Book a strategy call
Build & Run Propositions

2 · Build & Run Propositions

Hand us an outcome and we build the agentic product or journey end to end — to the same control standards as our Clean-Room Demonstration: every agent action authorised, evidenced, revocable. On-device models and data that never leaves the building where it must not. A fixed, shippable scope.

  • Customer outcomes built from the journey backwards
  • Agents that act under explicit authority, evidenced end to end
  • On-device models; data that never leaves the building
Book a strategy call
Owned Autonomy Transformation

3 · Owned Autonomy Transformation

The whole shift: making the bank agentic-ready — accountable and sovereign — on a Capability·Context·Consent (C3) architecture, with the agent wired to regulated settlement (ISO 20022, tokenised deposits, agentic payments). Led from inside the bank, not from a deck.

  • Enterprise & strategic architecture (BIAN, TOGAF, SAFe), composable by design
  • Operating model and transformation run from inside the bank
  • Programmable and agentic payments — the agent connected to real money
See our approach

From idea to embedded capability

Discover

Discover

A short, senior-led engagement to map your goals, data constraints and where AI creates real leverage — and where it doesn't.

Pilot

Pilot

A focused build that proves value in weeks, not quarters — with evals, guardrails and a clear line back to the outcome you care about.

Scale

Scale

We embed teams, platforms and practices so the capability — and the confidence to use it — stays with you when we leave.

Built by squads, not committees

We organise delivery into small, senior squads — each owning the Domain Building Blocks it ships, grouped into SAFe value streams — then reshape the wider bank around them, from risk and finance to regulatory affairs. An opinionated model on proven frameworks (BIAN, SAFe, TOGAF, SFIA). Roles are defined against SFIA — built by a SFIA 10 contributor.

1Product Owner
1Forward Platform Engineer
2Engineers
1Data Engineer

Our Clean-Room Demonstration

This is what accountable banking looks like, not slideware. An interactive demonstration of differentiated banking propositions on one governed, agent-native platform, built to production-grade controls on synthetic data. It demonstrates the technical capability: the model behind it is production-quality in design, though the platform is not yet product-ready.

Three propositions — Family Wealth, International Student and Premium Yacht — run on the live engine; Franchise Network is honestly labelled a blueprint. Click through each, watch a governed agent prepare and execute a journey, and see authority, consent and evidence kept explicit at every step.

Explore the demonstration

What we're learning, in the open

Field notes on agentic engineering, on-device AI and transformation — written from the work, not the brochure.

Leadership

Sailesh Panchal

Sailesh Panchal — Founder & CEO

A people and technology leader with 30+ years in financial services and FinTech. Today, building agentic product squads and production AI platforms for organisations that handle sensitive, regulated data.

Director of Engineering, Metro Bank — led a 1,600+ FTE function through digital transformation with Agile/SAFe value-stream teams. CTO, Orwell Group — the first UK payments system (Faster Payments, CHAPS, SWIFT, SEPA) and distributed ledger certified for cloud by regulators in the UK and France (FCA/ACPR), winner of the 2018 FIA Award.

Chief Architect, New Payments Architecture (PSR). Head of Regulatory Architecture, Lloyds Banking Group — technology sponsor for PSD1, PSD2, 5AMLD, GDPR, CRS, CASS, payM, IT resilience and cyber defence. CTO, Travelex. Chief Architect (ETC), JPMorgan. Currently writing a book on digital transformation in UK banking.

Our Partners

GFT JCurve BizzDesign

Our Sister Companies

Our Accreditations

FSQS Accreditation

Let's build something that ships.

Tell us what you're trying to do with AI. We'll tell you, honestly, whether and how we can help.

Book a call

Get the insights by email

Field notes on agentic engineering, on-device AI and transformation. No spam, unsubscribe anytime.