What we do · 02 · Technology & Intelligence
Software, AI and infrastructure run as one platform.
We build and run the technical foundation a brand needs — software, AI systems, infrastructure, and the security and support to keep it all dependable.
- Capabilities
- 10
- Scope
- Discovery → run
- Delivery
- AI-native
A live topology of Your platform in four layers: Experience (Websites & Apps, Search & AI Visibility); Intelligence (AI Strategy & Agents, AI Product & Automation); Platform & data (Custom Software & Data Platforms, AI Infrastructure & Cloud, Integration & Support); Trust & operations (Cybersecurity & AI Trust, Audits & Assessments, Tech Workforce). Requests travel between the capabilities while an illustrative status bar reports p95 latency of 212 milliseconds, 148 of 148 evals passing, 71 percent of the error budget left and 0.18 grams of CO2e per request.
Why now
Technology is no longer a website and a server. It is software, data and AI that must stay fast, secure and affordable every day.
Where the stack runs nowYour cloud account, your region, observable end to end
-
→AI in the product
LLM01
Prompt injection, first on the list
AI moved into the product
Features, agents and copilots now sit in customer journeys, so they bring a new class of bug that ordinary testing does not catch.
Source · OWASP Top 10 for LLM Applications, 2025
-
→1M+ tokens
1M+
Tokens of context, up from about 4k
Models change every month
Context windows, prices and rankings move faster than release plans. Evals that run on every change replace one-off testing.
Source · Provider model documentation, 2023–2025
-
→1 cited answer
10 → 1
Links became one answer with sources
Answers replaced links
AI Overviews, ChatGPT search and Perplexity answer first and cite a few sources. Being one of them is now a technical job.
Source · Google, OpenAI and Perplexity product pages
-
→Hours to report
6 h
To report a cyber incident in India
Scrutiny rose
The DPDP Act 2023, the EU AI Act and CERT-In six-hour incident reporting make security, privacy and AI governance delivery work.
Source · CERT-In Directions, 28 April 2022
The platform
Ten capabilities, one architecture.
Every capability builds a layer of the same system. Pick a layer to see who builds it, or follow one customer question through all four.
An architecture diagram in four layers. Experience: website and web app, mobile app, support chat, search and AI answers. Intelligence: support agent, retrieval, workflow automation, evals. Platform and data: custom apps and CRM, customer data, integration bus, AI gateway, GPU and serverless compute. Trust and operations runs alongside every layer: security and guardrails, observability, audits and the on-call squad. The steps below trace one request through it.
Builds the whole platform
Ten capability pages, four layers
L1 · Experience
L2 · Intelligence
L3 · Platform & data
L4 · Trust & operations
L1 · built by 2 capabilities
Experience
What customers and search engines touch: web, mobile and answer surfaces.
L2 · built by 2 capabilities
Intelligence
Agents, retrieval and automation that do work inside the product.
L3 · built by 3 capabilities
Platform & data
Custom apps, customer data, the integration bus, the AI gateway and compute.
L4 · built by 3 capabilities
Trust & operations
Security, audits and the people who run it, touching every layer.
One request, traced Illustrative
- 01 A customer asks the support assistant on the website. 0 ms
- 02 The AI gateway routes the question to a small model first. 38 ms
- 03 Retrieval reads the customer profile from the CDP through the integration bus. 96 ms
- 04 Guardrails check the draft answer before it leaves. 131 ms
- 05 The answer streams back and the trace lands in observability. 212 ms
- 06 Audits sample traces like this one every month. sampled
- 07 The squad on call owns it when anything drifts. owned
Platform Composer
Pick an outcome. See the platform it needs.
Choose a goal and the composer lights the capabilities involved, draws the order we would build them in, and lays out the phases, the team, the gates that must pass and what we measure in the first 90 days.
Plan: Launch an AI support assistant · 6 phases · 12 weeks · 6 capabilities
Architecture · capabilities involved6 of 10
The capabilities involved in the selected plan are highlighted on the platform map and numbered in build order.
Plan 1 of 5 · typical
Launch an AI support assistant
Answers customer questions from your own knowledge and records, hands off to people when unsure.
- Weeks
- 12
- Phases
- 6
- Capabilities
- 6
- 03 Use-case & risk assessmentAI Strategy Wk 1–2
- 02 Knowledge & CRM data readyPlatforms Wk 2–4
- 04 RAG assistant with evalsAI Product Wk 3–8
- 05 Gateway, caching & cost capsAI Infrastructure Wk 6–9
- 06 Red-team & OWASP LLM Top 10 reviewSecurity Wk 7–8
- 07 Human handoff & ongoing supportIntegration Wk 9–12 →
Team
- AI lead
- Product designer
- 2 × AI engineers
- Data engineer
- Security engineer
- Your support-operations lead
Quality gates · must pass
- ✓Faithfulness ≥ 0.90 on the golden set
- ✓Zero critical red-team findings
- ✓p95 < 2.5 s, first token < 800 ms
- ✓Handoff to a person on low confidence
- OWASP Top 10 for LLM Applications — LLM and generative AI security risks
- NIST AI RMF 1.0 — AI Risk Management Framework
- ISO/IEC 42001:2023 — Artificial intelligence management systems
- DPDP Act 2023 — Digital Personal Data Protection Act, 2023
First 90 days · we measure
- 01Containment rate
- 02Cost per resolved conversation
- 03CSAT on assisted conversations
Plan 2 of 5 · typical
Replace a legacy CRM
One customer record your sales and service teams trust, migrated without losing a quarter.
- Weeks
- 18
- Phases
- 6
- Capabilities
- 6
- 09 Data & process auditAudits Wk 1–3
- 02 Data model & CRM buildPlatforms Wk 3–14
- 01 Sales & service interfaceWeb & Apps Wk 5–14
- 07 ERP, email & telephony integrationsIntegration Wk 6–14
- 06 SSO, access model & audit loggingSecurity Wk 8–12
- 10 Cutover squad & hypercareWorkforce Wk 14–18
Team
- Solution architect
- 3 × full-stack engineers
- Data engineer
- Integration engineer
- QA engineer
- Your sales-operations owner
Quality gates · must pass
- ✓100% row-level reconciliation of migrated records
- ✓Two parallel-run cycles, zero P1 defects
- ✓Role-based access reviewed and signed off
- ✓Rollback rehearsed before cutover
- ISO/IEC 27001:2022 — Information security management systems
- SOC 2 — Trust Services Criteria
- DPDP Act 2023 — Digital Personal Data Protection Act, 2023
- OWASP ASVS — Application Security Verification Standard
First 90 days · we measure
- 01Time from lead to quote
- 02Duplicate record rate
- 03Weekly active users
Plan 3 of 5 · typical
Make the product fast on real phones
Good Core Web Vitals in the field, on the mid-range Android phones most customers actually carry.
- Weeks
- 10
- Phases
- 6
- Capabilities
- 5
- 09 Field-data performance auditAudits Wk 1–2
- 01 Image, font & JavaScript budgetsWeb & Apps Wk 2–6
- 05 Edge caching & CDN rulesAI Infrastructure Wk 3–6
- 01 Render path & hydration fixesWeb & Apps Wk 4–9
- 08 Template-level search checksSearch Wk 6–9
- 07 Budgets in CI & real-user monitoringIntegration Wk 8–10 →
Team
- Performance lead
- 2 × frontend engineers
- Platform engineer
- Search specialist
Quality gates · must pass
- ✓LCP ≤ 2.5 s at p75, field data
- ✓INP ≤ 200 ms at p75
- ✓CLS ≤ 0.1 at p75
- ✓JavaScript budget enforced on every merge
- Core Web Vitals — Loading, interactivity and visual stability
- WCAG 2.2 AA — Web Content Accessibility Guidelines
- SCI · ISO/IEC 21031:2024 — Software Carbon Intensity
First 90 days · we measure
- 01p75 LCP on low-end Android
- 02Mobile conversion rate
- 03Organic sessions on mobile
Plan 4 of 5 · typical
Get cited in AI answers
Your pages ranked on Google and quoted by AI Overviews, ChatGPT and Perplexity for the questions buyers ask.
- Weeks
- 12
- Phases
- 6
- Capabilities
- 5
- 09 Search & answer visibility auditAudits Wk 1–2
- 08 Entity & prompt mapSearch Wk 2–4
- 01 Technical SEO & structured dataWeb & Apps Wk 3–7
- 08 Answer-ready content systemSearch Wk 4–12
- 02 Product & knowledge data feedsPlatforms Wk 5–9
- 07 Prompt-panel tracking & reportingIntegration Wk 8–12 →
Team
- Search lead
- Content strategist
- Technical SEO engineer
- Data engineer
- Analyst
Quality gates · must pass
- ✓Valid structured data on every key template
- ✓Core Web Vitals good at p75
- ✓Facts consistent across site, feeds and profiles
- ✓Prompt-panel baseline recorded before changes
- Core Web Vitals — Loading, interactivity and visual stability
- WCAG 2.2 AA — Web Content Accessibility Guidelines
- GDPR — General Data Protection Regulation (EU) 2016/679
First 90 days · we measure
- 01Share of answer across a fixed prompt panel
- 02Citations per engine
- 03Organic and AI-referred conversions
Plan 5 of 5 · typical
Stand up an embedded AI squad
Vetted AI engineers inside your sprints, shipping a first feature to production with evals from day one.
- Weeks
- 12
- Phases
- 6
- Capabilities
- 5
- 10 Role profiles & squad matchWorkforce Wk 1–2
- 03 Use-case backlog & prioritiesAI Strategy Wk 1–3
- 06 Access, data handling & AI policySecurity Wk 2–3
- 10 Onboarding into your sprintsWorkforce Wk 3–4
- 04 First AI feature in productionAI Product Wk 4–10
- 09 Quarterly delivery & AI-readiness reviewAudits Wk 12 →
Team
- Squad lead
- 2 × AI engineers
- ML & data engineer
- QA engineer, eval focus
- Your product owner
Quality gates · must pass
- ✓Background checks and NDAs complete
- ✓Least-privilege access in your tools
- ✓Eval suite in CI before the first release
- ✓DORA metrics reported from sprint 1
- ISO/IEC 27001:2022 — Information security management systems
- ISO/IEC 42001:2023 — Artificial intelligence management systems
- DORA metrics — Software delivery performance
First 90 days · we measure
- 01Lead time for changes
- 02Change failure rate
- 03AI features shipped per quarter
Plans show typical phases and lengths from comparable programmes, not a quote. Every engagement is scoped with your team before it starts.
Let’s build what happens next.
Tell us what you’re building. We’ll answer straight.