Why safety software for the frontier, and why now.
Our thesis, the market as public figures describe it, what exists today, what we plan next, and what could go wrong.
Forward-looking statements. This page contains plans, targets and expectations about markets, products and the company. They are not promises. Actual results may differ, sometimes a lot. Market figures come from the public sources linked beside them, which we have not audited. Nothing on this page is an offer to sell, or a solicitation of an offer to buy, any security.
Three shifts are happening at once on India's frontiers.
01
The line is being instrumented, by many vendors at once.
In May 2026 the Home Minister said a grid of drones, radars and cameras would cover nearly 6,000 km of the Pakistan and Bangladesh borders within a year, with pilots at 7 to 8 sites (ETV Bharat, The Tribune). The services are building a joint counter-drone grid on their air defence centres (Unmanned Airspace). Every one of these programmes buys sensors from several makers. Each needs fusion, alerting and an audit trail that work across them.
Hazards at the line are a daily risk, and public forecast data is underused.
Assam's 2026 floods killed 100 people by 10 August (Mongabay India). Himachal's monsoon killed 267, 159 of them in road accidents (Rozana Spokesman, 5 Sep 2026). Global river, weather and sea models are open, but few are turned into a decision for a boat crew, a road gang or a patrol. Safety products sell to disaster authorities as well as forces, and their free public versions earn trust.
Oversight is becoming a requirement, and we treat it as the product's core.
The core duties of the data protection rules apply from 13 May 2027 (Hogan Lovells Cadwalader). The draft Defence Acquisition Procedure 2026 asks vendors in its indigenous categories to own their source code, and counts a vendor as Indian only if resident Indians own and control more than half of it (draft DAP 2026). The Supreme Court has agreed to hear a challenge to police facial recognition (LiveLaw). Products that arrive with purpose binding, retention and a verifiable audit trail are easier to buy and easier to defend.
Our products
Sakshi in every deployment; Darpan and the public advisories as the public face.
Public figures that size the opportunity.
Defence capital
₹2.19 lakh cr
capital expenditure in the 2026–27 defence budget, within a total of ₹7.85 lakh crore.
iDEX's DISC 14 alone carried 82 problem statements in 2026 (iDEX). Several match our products, such as a composite situational awareness picture, counter-drone detection, and a weather decision tool for Himalayan passes.
A Rajya Sabha committee asked in August 2026 for one interoperable early-warning platform, with warnings reaching villages and wards instead of stopping at the district (Report 261).
The CAG found 59% of the Central Water Commission's existing flood telemetry stations not working (CAG, 2017).
Landslide bulletins run operationally in 3 of 21 districts and none in Ladakh, Jammu and Kashmir or Arunachal Pradesh (PIB, Aug 2025).
Drone interception on Punjab's border rose from about 3% to 55%, and counter-drone technology passes through a certification committee chaired by the DG BSF (MHA, Dec 2024).
Who buys, and how we charge.
Buyers we are building for
Border guarding forces and their frontier headquarters, through modernisation budgets and the Smart Border pilots.
The armed forces, through iDEX and ADITI challenges that match a product.
State and district disaster management authorities, for flood, heat, landslide and lake-burst risk.
State police anti-drone units, fisheries departments, road builders and hydropower operators.
These are target buyers. We do not name customers on this site.
How we charge
A paid pilot of five to six months on one stretch of river, coast or border, billed at the end.
Then an annual subscription per site, or per kilometre of coverage, with updates and support.
Prahari pods sold with a maintenance contract.
Development grants where a government challenge fits the roadmap.
Public versions stay free. That is a doctrine commitment, not a funnel.
What exists on 21 October 2026.
Live
Seven products, counting the console, reviewed and running as a public demo on public data, with simulated parts labelled. Each has a manifest, tests and a methods note.
Built and tested
The Sakshi audit chain library with tamper tests, and an access policy with published test vectors.
In build
Fifteen more products in build. Each goes live only after review, and its status is public on the products page.
Product milestones.Forward-looking. Dates are targets, not commitments.
Q4 2026
Every product's first version on real or clearly labelled simulated data, with a manifest, tests and a methods note. Person profiles and scores gone from the console. Sandhi tracking Farakka flows against the treaty schedule before its term ends on 12 December 2026. Startup and procurement registrations done.
Q1 2027
A first flood pilot with one state or district authority, and a fisheries proposal. Cold-stress, winter-road and fog outlooks for the winter. A Prahari bench pod. An iDEX application where a challenge fits.
Q2 2027
Flood thresholds reviewed with the pilot authority before the monsoon. Data protection controls live before the core duties apply on 13 May 2027. A first Sangam adapter for one sensor maker's feed. Prahari's first river trial, at a non-border site.
Q3 2027
Monsoon operations with a daily refresh and an on-call rota for the flood, road, lake and alert products. Prahari's monsoon survivability test. A fisheries pilot on one coast. An air-gapped deployment kit and an independent review of the audit chain.
Q4 2027
Post-monsoon reviews and a second state for the flood products. Sangam SDK 1.0 with fusion metrics on public scenarios. The second transparency report.
What could go wrong.
Procurement
Government buying is slow and uneven, and pilots may not convert. The 2026 acquisition procedure was still a draft in October 2026 (analysis, 14 Oct 2026).
Data terms
Several products use Open-Meteo, whose free API is for non-commercial use. Commercial deployments need a subscription or a self-hosted instance, which we budget for.
Hardware
Prahari needs capital and long field trials. Tides, silt and salt are unforgiving.
Misuse
Our users hold power over people. The doctrine, Sakshi and published methods are our answer, and we expect to lose deals that would need us to cross a red line.
Policy
Treaties, border cooperation and data rules change. Products are built to stay useful when they do.
Ownership
We intend to stay majority owned and controlled by resident Indian citizens and to keep all source code owned in India, as the draft procedure expects of Indian vendors. This is a statement of intent.