Professional services

Engineers who stay for the go-live.

Software is the easy half. We take responsibility for the integrations, the cutover night, the operator training and the years afterwards — with named engineers, not a ticket queue.

Service lines

Six ways we work with you.

Engage us for a single assessment or for the whole programme. Every engagement produces documentation your team owns.

01 · Implementation

Product implementation

Installing and configuring our platforms in your environment — sizing, hardening, high availability, content design and operator enablement.

  • Environment sizing & architecture document
  • Installation, hardening & HA configuration
  • Content, layout & rule design workshops
  • UAT scripts, cutover plan & rollback path
Flight display platform
02 · Integration

Systems integration & migration

Connecting our platforms — or yours — to the systems that already run your operation, and moving data off the tools you're retiring.

  • AODB, PSS/DCS, GDS, ERP & access-control interfaces
  • Legacy VMS and FIDS migration with parallel running
  • Event bus, API gateway & message contracts
  • Data extraction, cleansing & reconciliation
ERP & enterprise suite
03 · Applied AI

AI & agent engineering

Building agentic systems on Claude models for your workflows — with the evaluation harness and guardrails that make them safe to put in front of customers.

  • Use-case assessment & feasibility spike
  • Tool design, retrieval & policy grounding
  • Evaluation suites, red-teaming & regression gates
  • Cost, latency & token-budget engineering
Disruption management agents
04 · Computer vision

Vision deployment & tuning

Turning an existing camera estate into reliable analytics — the unglamorous work of angles, lighting, thresholds and false-alarm reduction.

  • Camera survey, placement & lens recommendations
  • Model selection, fine-tuning on your footage
  • Rule tuning to target false-alarm rates
  • Edge GPU sizing & bandwidth planning
Vision VMS & analytics
05 · Managed operations

Managed operations

We run the platform for you — monitoring, patching, capacity and incident response — while your team runs the airport, airline or plant.

  • 24×7 monitoring with on-call escalation
  • Patching, upgrades & certificate lifecycle
  • Backup, restore drills & DR testing
  • Monthly service review with metrics
06 · Enablement

Training & enablement

Your staff should not need us for day-to-day changes. We train operators, administrators and developers, and leave the material behind.

  • Operator and supervisor certification sessions
  • Administrator & integration developer training
  • Runbooks, SOPs & recorded walkthroughs
  • Handover to internal support with shadowing

How to buy

Engagement models.

Pick the commercial shape that fits your procurement process — we will work inside any of them, and we will tell you which one suits the risk profile.

  1. Model A

    Fixed-scope project

    A defined deliverable, milestone payments and a written acceptance test. Best when the integration surface is known — a FIDS rollout, a VMS migration, an ERP module.

  2. Model B

    Paid pilot, then scale

    Four to six weeks on your real data in one terminal, one camera cluster or one route. A fixed pilot fee, a go/no-go review, and pilot cost credited against the rollout.

  3. Model C

    Dedicated engineering pod

    A named team — engineers, an architect and a delivery lead — reserved for you monthly. Suits evolving AI programmes where scope is discovered as you learn.

  4. Model D

    Outcome-linked

    Part of the fee tied to an agreed operational metric: display uptime, false-alarm rate, containment rate on concierge conversations, or minutes saved per disruption event.

Support & SLA

Three support tiers.

Every licence includes Standard. Mission-critical deployments — anything a passenger sees — should be on Critical.

Standard

Included with licence

For back-office and non-passenger-facing deployments.

  • Business hours, 09:00–18:00 IST, Mon–Fri
  • P1 response within 8 business hours
  • Email & portal ticketing
  • Quarterly maintenance releases
  • Documentation & knowledge base
Most chosen

Enhanced

18% of licence value / year

For multi-site production deployments with an internal ops team.

  • Extended hours, 07:00–23:00 IST, 7 days
  • P1 response within 2 hours
  • Named support engineer & escalation path
  • Proactive health monitoring & monthly report
  • Priority bug fixes and configuration assistance
  • Two upgrade assists per year
Discuss Enhanced

Critical

24% of licence value / year

For passenger-facing and safety-adjacent systems.

  • 24×7×365 with on-call rotation
  • P1 response within 30 minutes
  • Contractual restoration targets & credits
  • Quarterly DR drill and capacity review
  • On-site attendance for P1 where required
  • Annual architecture review with our CTO

Before you ask

Questions we get in every first call.

Can you deploy fully on-premise or air-gapped?

Yes for FIDS, Vision VMS and ERP — those run entirely inside your network with no outbound dependency. The AI agent products need model access, which can run through your own Anthropic API account, AWS Bedrock in your VPC, or a private endpoint; a fully air-gapped agent deployment is not something we will promise, because the model has to be reachable.

Do you replace our existing VMS or work alongside it?

Both are viable. We regularly run our analytics engine against an incumbent VMS's streams so you keep your recording estate, and migrate the VMS later if the case is there. Parallel running for a defined period is part of the standard migration plan.

Who owns the code and the data?

You own your data, always, including exports in open formats on exit. Platform code and our pre-existing IP remain ours and are licensed to you; custom integrations and configuration built specifically for you are assigned to you. Source escrow is available for on-premise deployments.

How do you price AI usage when volumes are unpredictable?

Model inference is passed through at cost with a stated margin, on a monthly statement, and we set hard token budgets per workflow with alerting. During the pilot we measure real cost per conversation or per disruption event so the rollout price is based on your data rather than an estimate.

What happens if the pilot fails?

It ends. You keep the assessment, the architecture document and the measured results, and there is no obligation to proceed. We would rather lose a pilot fee than carry a reference customer who was pushed into a rollout that could not work.

How large is the team behind this?

We are a small, senior engineering company — the people who build the products are the people who show up to your integration workshops. For large rollouts we staff installation and commissioning through vetted local partners under our supervision, and we will name them before you sign.

Let's build

Start with a scoping call.

Forty-five minutes with an engineer, not a salesperson. Bring your systems list and your constraints; you'll leave with an honest view of effort and sequence.