Engineering depth held to global standards — without the marketing claims.
We restructured the firm around AI/ML because the market gave buyers two options: under-engineered
offerings at low cost, or capable systems priced for global rather than regional operations.
Neither served the organisations actually running schools, hospitals, plants, VFX studios, or
production teams.
We focused the firm on a third option — engineering depth at a deployment cost that fits
operational reality, applied to whichever AI/ML problem the client needs solved. Today, we
operate across five practices.
Placeholder image
What we do.
We build production AI/ML systems across five practices:
Safety & computer vision — our lead practice. Multi-site CV platforms for schools, hospitals, plants, retail, and corporate campuses.
Agentic systems — orchestration, tool use, MCP integration, with observable production behaviour.
VFX & animation pipeline — generative tooling for Nuke, Houdini, Maya, and Blender.
AI-native film production — pre-production through distribution, with IP and consent guardrails.
Custom ML & research — the bespoke work that doesn't fit any of the above.
Every practice commits to written measurement and a published "what we won't build" list.
The team.
A small, senior team of CV and ML engineers, agent-systems engineers, generative-media
specialists, SREs, and integration architects. Deliberately cross-practice: the engineers
designing the edge-inference platform also understand what production agents need, and the
same supervisors reviewing VFX pipelines understand the colour pipeline of film delivery.
We are deliberately small. We sell engineering capacity, not headcount.
Where we are.
A distributed engineering team across Pune, Hyderabad, Bengaluru, and Delhi-NCR — with the
registered office in Pune, Maharashtra. We deploy across India, with
client work concentrated in Maharashtra, Karnataka, Tamil Nadu, Telangana, and Delhi-NCR. Public-sector
engagements are handled separately under appropriate clearances.
Values
The commitments we hold under commercial pressure.
These are not aspirational statements. They are the boundaries we have declined to cross in pursuit of revenue, and the boundaries we will continue to hold across every engagement.
Child safety as the primary consideration
K-12 deployments are the highest-stakes setting in our portfolio. Children's biometric data receives dedicated engineering attention, a separate accuracy SLA structure, and a privacy posture that is not negotiable on commercial grounds.
Defined responsibility boundaries
What we will not build is as material as what we will. The categories we decline — emotion recognition, predictive misconduct scoring, behavioural flagging in proctoring — protect both buyer and bystander, and are documented in each engagement contract.
Transparent commercial terms
Quotes are built bottom-up, without discount margin pre-loaded for negotiation. The figure we present in Discovery closely tracks the figure we deliver against. Commercial terms are confirmed during the engagement, not on the website.
Contracted, measurable SLAs
Performance figures we are confident in delivering. We measure monthly. Service credits apply when we miss. We'd rather contract to a 90% target we meet than a 99% target that needs redefinition later.
Start a conversation with the engineering team.
Share the operational context, existing infrastructure, and compliance requirements. We respond within one working day.