Services

Five practices. One careful standard.

We take on work in these five areas. Most engagements touch two or three of them, because problems rarely respect the categories.

01

Agentic AI

Agents that hold a tool belt, know when to stop, and can be audited afterwards.

What we build

  • Multi-tool agents wired to your real systems: CRM, ledger, ticketing, warehouse.
  • Orchestration across steps — planning, retries, human approval gates on anything irreversible.
  • Retrieval over your own corpus with citations, not a general chatbot with a company name on it.
  • Guardrails: allow-lists, spend caps, refusal policies, and an eval suite that runs on every prompt change.

A pilot could look like

A focused accounts-payable pilot could read a sample invoice set, extract line items, match purchase orders, and route every uncertain decision to a human before anything is posted.

Stack

  • TypeScript
  • Vercel AI SDK
  • Gemini / GPT
  • pgvector
  • Temporal
  • Postgres
  • OpenTelemetry

Delivery standard

Every action is logged. Irreversible steps require an agreed approval gate.

02

ML / Data Science

Models trained on your history, evaluated against the decision they actually inform.

What we build

  • Demand and revenue forecasting with backtests you can read.
  • Propensity, churn and risk scoring wired into the tools your team already opens.
  • Custom model training and fine-tuning where an off-the-shelf API genuinely isn't enough.
  • Monitoring for drift, plus a retrain schedule that someone owns.

A pilot could look like

A focused forecasting pilot could use a representative product set, compare a model against a simple baseline, and deliver recommendations in the tool your team already uses.

Stack

  • Python
  • PyTorch
  • scikit-learn
  • LightGBM
  • MLflow
  • Airflow
  • DuckDB

Delivery standard

A model must beat an agreed baseline on held-out data before it earns a production path.

03

Data Analysis

One number, one definition, one place to find it.

What we build

  • Warehouse modelling: raw sources to tested, documented marts.
  • Pipelines with alerting, so a silent failure isn't discovered at a board meeting.
  • Dashboards built around the four decisions a team actually makes each week.
  • Ad-hoc analysis when the question is one-off — we won't sell you a platform for it.

A pilot could look like

A focused data pilot could reconcile one decision-critical metric across your main sources, document its definition, and expose every transformation for review.

Stack

  • dbt
  • BigQuery
  • Postgres
  • Dagster
  • Metabase
  • Evidence
  • SQL

Delivery standard

Every reported number has an owner, a definition, a source, and a freshness check.

04

Web Development

Product engineering. Type-safe, tested, and fast on a bad connection.

What we build

  • Customer-facing products and internal tools, from first commit to production.
  • API and integration work: webhooks, auth, third-party systems that document themselves badly.
  • Performance and accessibility as build criteria, not a cleanup ticket.
  • CI, preview environments, and infrastructure you can hand to another team.

A pilot could look like

A focused product pilot could cover one complete workflow—sign-in, a core task, status tracking and administration—using real constraints rather than polished mock data.

Stack

  • React
  • TypeScript
  • TanStack Start
  • Tailwind
  • Postgres
  • Cloudflare
  • Vitest

Delivery standard

Acceptance criteria include accessibility, performance, tests, deployment and handover.

05

E-commerce

The storefront is the easy half. We do the other half too.

What we build

  • Headless storefronts and custom checkout flows where the default one costs you money.
  • Catalogue, pricing and inventory sync across marketplaces and the warehouse.
  • Post-purchase automation: delivery notifications, returns, review requests, fraud rules.
  • Analytics that tie spend to contribution margin rather than last-click revenue.

A pilot could look like

A focused commerce pilot could validate one high-friction journey, such as bundle checkout or order-status automation, before changing the wider storefront.

Stack

  • Shopify Hydrogen
  • Medusa
  • Stripe
  • Twilio
  • Klaviyo
  • Postgres
  • Next.js

Delivery standard

Changes are measured against a baseline and can be rolled back without disrupting operations.