Karis designs, builds and runs the intelligent systems that India’s cities, railways, airports and public institutions depend on, at the scale India needs.
Pause motion
Urban digital twin
Aerial intelligence
Fare collection and transit
Gate Pilot
Public systems
Operant
OpenSlate
Computer vision
Applied ML
For government
For PSUs
For enterprise
Urban digital twin
Aerial intelligence
Fare collection and transit
Gate Pilot
Public systems
Operant
OpenSlate
Computer vision
Applied ML
For government
For PSUs
For enterprise
About Karis
Karis Partners builds intelligence for the systems India runs on: digital twins for its cities, eyes in the air over its forests, software for its railways, metros and airports, and the upkeep of the public systems its people depend on.
Our vision: lasting partnerships, where human judgement and artificial intelligence build progress that endures.
How Karis delivers
01 Survey
02 Design
03 Build
04 Run
Survey: We start where the work happens. On the platform, at the gate, in the control room. We learn how a system really runs before we change it.
Design: We design for India’s scale. Peak hours, festival crowds, monsoon days, many languages. We plan for the hardest day, not the average one.
Build: We build into what already runs. No shutdowns and no rip-and-replace. New capability goes in beside the systems people already trust.
Run: And we stay until it holds. We operate, maintain and improve what we build, year after year.
That is technology, applied. Calm, exact and built to last. That is what we mean by grace.
We start where the workhappens.
On the platform, at the gate, in the control room. We learn how a system really runs before we change it.
We design for India’sscale.
Peak hours, festival crowds, monsoon days, many languages. We plan for the hardest day, not the average one.
We build into whatalready runs.
No shutdowns and no rip-and-replace. New capability goes in beside the systems people already trust.
And we stay until itholds.
We operate, maintain and improve what we build, year after year.
That is technology,applied.
Calm, exact and built to last. That is what we mean by grace.
Your teams show us the work as it really runs. You see where the friction is before anything is proposed.
You agree the design with us, and how its success will be measured, before any of it is built.
A narrow pilot on your real data first, measured. Then the full system, built and tested alongside your teams.
We stay to run it, measure it and extend it to the next problem.
01 · Survey
02 · Design
03 · Build
04 · Run
Stage 01 of 04 · Survey
[02]What we’re building right now
The systems a billion people will run on.
Programme 01/ 05
Cities · Urban planning
Urban digital twin
A living model of the Indian city.
A digital twin for the people who plan cities: land, roads, transit, utilities and population on one model. Run a scenario before a single brick moves, and see what it does to traffic, density and public services years ahead. The simulation and prediction are tuned for Indian cities, not imported from elsewhere.
Scenario simulation for roads, transit, zoning and utilities
Predictive analysis of growth, demand and load
Calibrated for the way Indian cities actually grow
Drone surveys turned into foresight. Over forests: canopy health, change over time and early warning of risk. Over cities: construction, encroachment and the condition of the built environment. Every survey is mapped and compared with the last, so action comes before the damage does.
Drone imagery processed into mapped, comparable layers
Predictive models for forest health and urban change
Annual maintenance contracts for state and national government authorities and public sector undertakings. Monitoring, upkeep, upgrades and response for the software public services depend on, held to the accountability that public work demands.
Continuous monitoring and preventive upkeep
Defined response and resolution commitments
Upgrades delivered without interrupting public service
Intelligent software for automatic fare collection, with an intelligence layer for Indian Railways and metro organisations. Ridership, gate throughput, revenue and anomalies read in one place, so operators can forecast crowds, reconcile revenue and run stations ahead of demand.
AFC software built for Indian rail and metro scale
Indoor navigation for airports. Gate Pilot guides passengers through the terminal, from entry and check-in through security to their gate, with live routes and walking times, inside buildings where GPS cannot reach.
Animation, an example: Operant groups repeated conversations into one pattern, learns a skill from them, and once the skill is approved every later run uses it; the cost per run falls from ₹19.30 to ₹7.60.
.env
# Point your agent here (was: api.anthropic.com)# Point your agent here (was: api.anthropic.com)ANTHROPIC_BASE_URL=https://app.operantlabs.comANTHROPIC_BASE_URL=https://app.operantlabs.comANTHROPIC_API_KEY=your-key-••••••••ANTHROPIC_API_KEY=your-key-••••••••
two lines · no code change
OpenSlate
The AI-native classroom. Any page of any book, made workable.
Any textbook becomes editable pages.
Lasso an equation: see it, simulate it, practise it.
Animation, an example: a scanned textbook page becomes editable text; the reader lassoes an equation, sees it as a 3D surface, runs gradient descent on it, and asks a question that is answered citing section 12.4.
skills
$ manim -qh SaddleFlow.py FieldEscape$ manim -qh SaddleFlow.py FieldEscape· gradient descent · η = 0.08 · 14 steps to escape· gradient descent · η = 0.08 · 14 steps to escape
a skill, rendered into the page
The studio
Every institution deserves a system built for it alone. We design yours, build it, and run it with you.
We begin where your work happens, beside the people who do it.
Every Karis capability, shaped to your data, your rules and your languages.
Animation, an example: Karis maps how an institution’s work moves today, fits four of its seven capabilities into a clean four-step flow, proves the system in one ward, station or district, then runs it across the whole city, network or state.
Karis Partners builds intelligence for the systems India runs on: digital twins for its cities, eyes in the air over its forests, software for its railways, metros and airports, and the upkeep of the public systems its people depend on.
Who do you work with?
Government bodies first, then enterprises, institutions and technical partners.
What are Operant and OpenSlate?
Operant sits between your agents and the model APIs they use. It observes real traffic, learns the repeated work and codifies it into approved skills, which cuts the cost of running AI. OpenSlate is an AI-native classroom: any page of any book, made workable.
How does an engagement start?
With a survey of the work as it really runs, beside the people who do it. Then design, then build (one narrow pilot on your real data first, measured, before anything scales), then run: we stay to operate and improve it.
