Kitsilva — AI Systems Engineering

01 / CAPABILITIES

Strategy & Advisory

Define a pragmatic data and AI strategy, then a plan to execute it.

Data Science & Analytics

Turn messy data into decisions, forecasts, and measurable insight.

Machine Learning Engineering

Design, train, and ship models that hold up in production.

AI Agents & LLM Solutions

Build agents and LLM systems that do useful, reliable work.

Data Engineering & Pipelines

Reliable pipelines and platforms your team can build on.

Operating-Model & Cost Transformation

Redesign how technology is organised, sourced and run — and take real cost out without breaking delivery.

M&A IT Due Diligence

Assess IT risk, cost and separation before a deal — and steer integration after it.

Fractional CTO

Senior technical leadership, on tap, without the full-time hire.

Program & Delivery

Scope, run, and land complex programmes end to end.

Bigger than a freelancer. Closer than a big firm.

Broader than a freelancer

A full team's range of capability, with the ability to scale up around a project.

Smaller than a big firm

Direct access to senior people — no layers, no inflated day rates.

Built on real operating experience

We've run commercial operations, startups, and enterprise programmes — not just advised on them.

For businesses that fall through the cracks

Serious technical partnership for organisations the large consultancies overlook.

AI PRICING SAAS

A machine-learning price-optimisation platform that moves prices to match value — reading customer-behaviour signals to find pricing headroom, evidencing the rationale behind each move, and pacing adjustments to avoid shocks.

Proprietary ML · value-based pricing · AWS serverless

Read case study →

AI trading system — LSTM selection and RL timing

QUANTITATIVE TRADING · AI ENGINEERING

A two-stage AI system combining LSTM share selection with reinforcement-learning timing — automated daily before the open, benchmarked rigorously against buy-and-hold.

+47% profit vs buy-and-hold · LSTM + RL · daily automation

Read case study →

AI AGENTS · EVENTS

Building the AI platform that turns chaos into intelligence at the world's largest trade shows — founding engineering partner from zero to production in three months.

65,000 users · 5 events · 100% API reliability at peak

Read case study →

DATA SCIENCE · MACHINE LEARNING

Predicting exhibitor retention from Customer Experience data — then quantifying what a 1% retention improvement is worth, event by event, to build a self-funding business case.

~80% accuracy · £2.8m EMEA opportunity · 85 events

Read case study →

Live revenue forecasting data visualisation

DATA SCIENCE · AI ENGINEERING

A live AI system that re-forecasts every trade show's revenue and space daily — so commercial teams always know where they'll finish, early enough to act.

3–4% median error · ~120 shows back-tested · daily pipeline

Read case study →

Boutique data & AI consulting that ships.

01 / CAPABILITIES

Strategy & Advisory

Data Science & Analytics

Machine Learning Engineering

AI Agents & LLM Solutions

Data Engineering & Pipelines

Operating-Model & Cost Transformation

M&A IT Due Diligence

Fractional CTO

Program & Delivery

Bigger than a freelancer. Closer than a big firm.

Broader than a freelancer

Smaller than a big firm

Built on real operating experience

For businesses that fall through the cracks

03 / SELECTED WORK

A machine-learning price-optimisation platform that moves prices to match value — reading customer-behaviour signals to find pricing headroom, evidencing the rationale behind each move, and pacing adjustments to avoid shocks.

A two-stage AI system combining LSTM share selection with reinforcement-learning timing — automated daily before the open, benchmarked rigorously against buy-and-hold.

Building the AI platform that turns chaos into intelligence at the world's largest trade shows — founding engineering partner from zero to production in three months.

Predicting exhibitor retention from Customer Experience data — then quantifying what a 1% retention improvement is worth, event by event, to build a self-funding business case.

A live AI system that re-forecasts every trade show's revenue and space daily — so commercial teams always know where they'll finish, early enough to act.

Chris Witheridge

Maksym Vierovkin

Have a problem in data, AI, or technology that doesn't fit a neat box?