Skip to main content

OUR WORK

Patterns we've owned. Outcomes we've sustained.

Evidence from the field. Not promises.

[ 001 / 009 ]

INDUSTRIES

Where intelligence complexity shows up first.

Financial Services & Fintech

Compliance, fraud detection, lending intelligence, customer experience.

Healthcare & Insurance

Claims automation, clinical decision support, patient engagement.

FMCG & Consumer Goods

Demand forecasting, supply chain intelligence, retail analytics.

Logistics & Supply Chain

Route optimisation, warehouse intelligence, real-time tracking.

Governance & Compliance

Policy automation, audit intelligence, regulatory monitoring.

Media & Entertainment

Content intelligence, audience analytics, rights management.

CASE STUDIES

What ownership looks like in practice.

IiAS
Fintech · Governance & ComplianceIndia · 12+ months · Ongoing

Rebuilding India’s proxy advisory infrastructure. Without taking it offline.

IiAS

~70%Voting workflow time reduction
100%SEBI stewardship reporting coverage
250Mutual fund houses served
1,250+Active compliance professionals
The Situation

IiAS is India’s leading proxy advisory firm. 250 mutual fund houses depend on them to research corporate resolutions, issue voting recommendations, and file SEBI-mandated stewardship reports. That’s…

What We Built
Unified data pipeline connecting all four products into one platform
Vote Management System built from the ground up
SEBI-compliant exports built into every workflow step, not bolted on after the fact
+2 more

This wasn’t a redesign. It was a rebuild of the infrastructure that India’s mutual fund industry relies on.

JoVE Coach
Education · STEM & Business LearningUS-based · Web Application

A fresh start. A full rebuild. Under one month.

JoVE Coach

1M+Learners in JoVE’s ecosystem
< 1 moFull rebuild time
0 → FullMobile responsiveness
70–80%Improvement in clarity scores
The Situation

JoVE is a B2B video library serving healthcare institutions, research labs, and universities. 1M+ learners across 1,800+ institutions. JoVE Coach was their entry into B2C: an AI-powered learning…

What We Built
Dual-mode interface: doubt mode for quick answers, microcourse mode for structured learning
Intent-capture nudges guiding casual browsers into course enrolment organically
Custom design system with non-standard UI elements built specifically for the platform
+2 more

The previous team treated every user the same. We rebuilt the product around how users actually think.

SortMyPrep
EdTech2.5 months · 2-person core team · Web Application

Six years of backlog. Permanently solved in 2.5 months.

SortMyPrep

2.5 moFull build time
2Person core team
6+ yrsContent backlog cleared
70–80%Teacher review time reduction
The Situation

SortMyPrep is a question bank for IB students: past papers, mark schemes, and detailed explanations. The demand was there. The market was clear. The problem was the production workflow: every…

What We Built
AI pipeline: Google Gemini generates step-by-step explanations for each question
Teacher reviews and refines via conversational interface. Guides the AI, doesn’t rewrite
Verified content publishes to students. Nothing reaches them without teacher sign-off
+2 more

The pipeline was the engine. The student features were the product. Both shipped in 2.5 months.

Arcmaze
Real Estate & Construction2–2.5 months · Web Application · Pilot

AI document search for a domain no model has ever seen.

Arcmaze

4–6h→2mQuery time reduction
70–85%Query routing precision
3–5 hrsSaved per architect per week
10–15Pilot users onboarded
The Situation

Construction drawings encode information in symbols, line weights, spatial relationships, and tiny annotations. No public dataset of annotated construction drawings exists. No AI model has been…

What We Built
Multimodal RAG platform with a query routing layer at its core
Auto-categorises documents by discipline at ingestion (mechanical, electrical, plumbing, structural)
Natural language query routing. Narrows retrieval context before the model processes anything
+2 more

We started with zero domain knowledge. Every design decision traces back to learning how architects actually think.

Benepower
Healthcare · InsurTechUS-based · Web + iOS + Android

We replaced a call centre with AI before anyone asked us to.

Benepower

100%Employee queries handled by AI
<10 secResponse time vs 24–48 hrs
70–80%Estimated support cost reduction
88–92%AI accuracy on real queries
The Situation

Every time an employee at a Benepower client had a health insurance question, a ticket went to an outsourced call centre. An agent would open the plan document, find the relevant section, and respond…

What We Built
AI knowledge base (chat + voice). Reads live plan documents, responds in under 10 seconds, 24/7
88–92% accuracy validated against real employee queries across multiple plan types
Employer portal: enrolment workflows, document management, third-party integrations
+2 more

Benepower hadn’t asked for this. We built the proof of concept anyway, validated it against real queries, and presented the case.

Have a perspective to discuss?

Pixeldust