Larger custom systems,
delivered.
Multi-service systems for problems that don't fit an off-the-shelf tool — execution control-towers, enterprise finance and sales analytics. Delivered by a small senior team as client engagements, with AI in the loop where it fits and rules where a decision must be auditable.
— Internal beta. Client work below is anonymized. Ankush to refine.
Spec-driven, senior, accountable.
A formal spec (BRD/PRD, provenance-tagged requirements), a small senior team owning both the engineering and the intelligence layer, and AI used through delivery — made explicit and auditable, not a black box.
Spec to build
Formal requirements with fact / inference / assumption tagging, so scope stays evidence-grounded and reviewable by engineering and business together.
One accountable team
The same senior team owns the backend and the AI layer — delivered inside your environment, not handed across silos.
Grounded AI
Where we use AI, it stays grounded and cited — never inventing numbers on a decision that a client acts on.
Anonymized client projects.
Real engagements, client identities removed. Outcomes are qualitative or client-stated targets where a project is still in build.
Execution control-tower for a global elevator & escalator OEM
Installation and modernization sites needed real-time visibility of execution status — job-card to completion — without replacing the core ERP. We built a control-tower system of record: Next.js + FastAPI, a state-machine domain engine, role-based access, and a Vitest/Playwright test suite. No AI — a deterministic operations system the client runs alongside SAP.
Conversational finance analytics on Microsoft Fabric
The same OEM wanted natural-language finance analytics without replacing its governed SAP + Power BI stack. We delivered a multi-agent layer on Microsoft Fabric: an AI Foundry orchestrator routing to a Fabric Data Agent over the certified semantic model, with a strict "never invent numbers, always trace to source" governance prompt and a large regression-tested query suite. Reached final-demo, integrated into Microsoft Teams.
Sales & inventory analytics agent
A MENA-region FMCG leader's sales and inventory data sat on SharePoint with no unified analytics. We delivered a Fabric Data Agent + Power BI semantic model behind a Copilot Studio sales agent — intent classification, certified-measure DAX queries, TOPN capping for high-cardinality data, role-based access, and an insight layer that describes only what the data shows. Delivered inside Microsoft Teams.
Anonymized — no client names, logos, or confidential figures. These are Flexilytics client engagements, not products; some are in build, so outcomes are qualitative or client-stated targets.
More anonymized case studiesOne team accountable for both the engineering and the intelligence layer.
Talk to us