SmartTerra is using AI to plug India’s massive water loss problem

Indian cities are bleeding water, relentlessly and at scale. Nearly half of all treated water never reaches a tap. Delhi loses about 52% of its supply, while Bengaluru loses roughly 37-40% as water leaks through cracked pipelines, corroded joints, and faulty valves. Cities invest crores of rupees to source, treat, and pump water, only to watch much of it leak away before anyone can use it. 

The consequences are visible every day. Most urban neighbourhoods receive water for just one to six hours daily, forcing residents to rely on storage tanks, diesel-powered tankers, illegal connections, and groundwater extraction to bridge the gap.

The technology to detect leaks already exists. IoT 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 In practice, that data is scattered and underused, spread across outdated maps, disconnected systems, and maintenance logs.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. The Bengaluru-based startup uses artificial intelligence (AI) to 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 it works

SmartTerra's AI and machine learning software simultaneously analyses multiple data sources, including network maps, maintenance records, flow meters, pressure sensors, and consumer meters. 

“Our algorithms spot patterns that signal trouble,” Santhanam explains. 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 (systems that monitor and control water infrastructure), 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-meter pipe stretch, check specific suspect meters, or investigate a junction showing pressure drops.

“We achieve up to 75% accuracy in pinpointing leaks and illegal connections, and 80% accuracy in detecting faulty meters,” Govindu says. The AI learns from field feedback, with future accuracy improving when crews confirm or correct predictions. 

"We've demonstrated measurable water loss reduction in as little as three months," Santhanam says. "This works with intermittent supply networks like in Indian cities and 24/7 systems elsewhere."

Focus on quality 

Municipalities are SmartTerra's main market, but the company now serves industries where water quality is critical. Many manufacturing processes are highly sensitive, and even minor contamination can ruin entire production runs.

"One of our clients manufactures mirror glass and needs water at 2 parts per million (ppm) TDS, far purer than drinking water (40-70 ppm),” Govindu says. “If the TDS rises even slightly, an entire day's output has to be scrapped.”

SmartTerra monitors water and sewage treatment plants in real time, tracking flow, pressure, and quality parameters like pH and TDS. When anomalies appear, the system alerts operators early, while its mobile app translates alerts into immediate field actions.

The technology behind SmartTerra 

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 and incomplete records. The mobile app works 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. The platform undergoes regular security testing, and consumer identifiers are excluded from the system.

Two models, one goal: Stop the leaks

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 Rs 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. 

SmartTerra 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. The startup currently has active contracts in 12 cities and 2 industrial campuses.

SmartTerra’s 30-member team works directly with municipalities through tenders and partners with engineering companies, treatment plant manufacturers, and sensor vendors to scale in Australia and the Middle East. According to the Business Research Company, the global water management systems market is expected to grow from $17.01 billion in 2025 to $29.29 billion by 2030, at a CAGR of 11.4%. SmartTerra competes in this space with players such as India’s Agua Wireless Systems and US-based Ketos Water. 

Backed by $1M to plug India's water crisis

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 is based on 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. 

Long-term, the company is piloting wastewater network analysis to address flooding and sewer management. “Our goal is to become an integrated water utilities solution, providing intelligence across water supply, wastewater, and stormwater systems," Santhanam says.

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.”