How SmartTerra Is Using AI to overcome India’s Massive Water Loss Problem
Water scarcity is a problem for many in India, but one such problem that exists but isn’t visible is how up to half of urban water is lost before reaching taps. Meet SmartTerra, which is using AI to pipeline data, predict leaks, and reduce water loss for cities and industries across India and Southeast Asia.
Indian cities are losing massive amounts of water; nearly half of the treated water never reaches your taps. Delhi loses about 52% of its supply, while Bengaluru regularly loses roughly 37–40%. Water escapes from leaking pipelines, corroded joints, and faulty valves. Urban India spends crores on water treatment to reduce water loss.
The consequences are visible every day; most urban neighbourhoods receive water for just one to six hours daily. A majority of the population relies on storage tanks, diesel-powered tankers, illegal connections, and groundwater extraction to bridge the gap. The technology to detect leaks already exists. A lot of sensors can monitor flow and pressure across vast pipeline networks, and smart meters track consumption in homes and businesses. In theory, cities have enough data to pinpoint where water is being lost. In practice, that data remains scattered and underused, spread across outdated maps, disconnected systems, and maintenance logs.
This gap between data and decision-making led to the founding of SmartTerra in 2017. This Bengaluru-based startup uses AI to monitor and analyse scattered water network data and predict leaks before they become a crisis.
The company is led by CEO Gokul Krishna Govindu, a former Apple executive; COO Giridharan Sengaiah, an urban water management specialist from PwC; and Chief Scientist Navaneethan Santhanam, an AI expert who previously worked at Freshworks.
How SmartTerra works using AI
SmartTerra uses AI and machine learning software simultaneously and analyses multiple data sources, including network maps, maintenance records, flow meters, pressure sensors, and consumer meters.
Chief Scientist Navaneethan Santhanam explains that “Our algorithms spot patterns that signal trouble”. This includes pipes likely to leak, abnormal flow behaviour indicating existing leaks, faulty meters under-registering consumption, and illegal connections siphoning off supply.
The platform integrates directly with utilities’ existing systems such as SCADA, billing software, and maintenance logs, bringing fragmented data into a single view. It then builds a digital model of the water network and flags issues with specific locations. Instead of vague alerts, field teams receive actionable instructions: inspect a 200-metre pipe stretch, check specific suspect meters, or investigate a junction showing pressure drops.
By turning scattered data into precise, location-specific insights, SmartTerra helps utilities tackle water loss that has long been treated as an unavoidable cost of urban supply.
Expanding Beyond Municipalities
Municipalities remain SmartTerra’s main market, but the company serves industries as well because if companies get contaminated water, the whole production gets affected.
SmartTerra monitors water and sewage treatment plants in real time, tracking flow, pressure, and quality parameters such as pH and TDS. When anomalies appear, the system alerts operators early, while its mobile app translates those alerts into immediate field actions.
The Technology Behind the Platform
At the core of SmartTerra’s platform is its proprietary data engine. Before any analysis, algorithms automatically clean incoming data, correcting outdated maps, accounting for failed sensors, and filling gaps from manual records.
“Without this step, analytics are useless,” Santhanam says. Unlike Western software built for continuous supply and clean data, SmartTerra is designed for India’s intermittent supply patterns and incomplete records. The mobile app is simple enough for field staff without technical training; public utility workers across Telangana, Tamil Nadu, and Karnataka use it daily.
Each client’s data is stored in isolated databases with cybersecurity and privacy controls. “We regularly undergo Vulnerability Assessment and Penetration Testing for cybersecurity certification,” Govindu says. Consumer identifiers are excluded from the system.
Two Engagement Models
SmartTerra operates across B2B and B2G through two engagement models. Its “stability model” targets cities or factories with high water losses, deploying the company’s teams for six to nine months to deliver specific outcomes such as cutting losses by 10–20 percentage points or improving network pressure. Once systems stabilise, clients move to the “continuity model,” a SaaS subscription that they run independently.
Onboarding takes three to four weeks to integrate existing data systems and configure the program. “From week five onward, the software begins generating actionable insights,” Santhanam says. “We often recommend additional sensors to fill data gaps.”
Pricing for the stability model ranges from ₹3–5 lakh per 10–15 km of network, while industrial pricing is based on the volume of water treated daily.
One of SmartTerra’s largest deployments was with L&T in Pune, where it delivered verified savings of 10.8 million litres per day across 150 km of network in nine months. The startup has run pilots in 17 cities across India and Southeast Asia, with ongoing trials in Australia. Several pilots have converted into paid contracts with clients including Suez, L&T, and industrial operators. SmartTerra currently has active contracts in 12 cities and two industrial campuses.
Funding and the Road Ahead
In 2025, SmartTerra raised $1 million, led by Siana Capital, which invested approximately $800,000. Additional investors included angels such as Shalini Chhabra and Kumar Ganapati from impact investment firm Three Eye Partners. The investment reflects demonstrated traction across India, Singapore, the Philippines, and Indonesia, with an expanding presence in the US, Asia-Pacific, and Gulf regions.
The company’s immediate focus is to stabilise its industrial water segment, launched in mid-2025, while pursuing scaled commercial contracts in the Middle East and Australia. SmartTerra aims to expand operations to 50 cities in the next three years.
SmartTerra is also eyeing rural expansion. “Each panchayat needs an extremely affordable digital solution,” Govindu says. “Rural expansion will take time, but it’s a core part of our vision.”
By converting fragmented data into precise, actionable intelligence, SmartTerra is helping cities and industries confront a problem that has long been accepted as inevitable. Reducing water loss is no longer only a matter of replacing pipes; it is increasingly a matter of making better use of the information that already exists.
FAQ
Q: How much water do Indian cities typically lose?
A: Nearly half of treated water never reaches taps. Delhi loses about 52%, while Bengaluru loses roughly 37–40%.
Q: How does SmartTerra detect leaks?
A: Its AI analyses network maps, maintenance records, flow meters, pressure sensors, and consumer meters to identify patterns that indicate leaks, faulty meters, or illegal connections.
Q: What accuracy does the system achieve?
A: Up to 75% accuracy in pinpointing leaks and illegal connections, and 80% accuracy in detecting faulty meters.
Q: Who are SmartTerra’s main customers?
A: Municipalities are the primary market, with growing adoption among industries that require high water quality.
Q: How quickly can results be seen?
A: The company has demonstrated measurable reductions in water loss in as little as three months.