ABOUT COGNICITY

Building technology with purpose.

An Indian deep-technology company engineering the intelligence powering tomorrow’s world.

Cognicity edge intelligence

Who we are.

Cognicity is an Indian deep-technology research and product engineering company, registered in April 2025. We develop indigenous technologies that enable reliable on-device intelligence for resource-constrained and mission-critical environments.

We specialize in intelligent embedded systems that can sense, process, analyse, communicate and make autonomous decisions at the point of operation, without depending on centralized computing infrastructure.

Our research and product engineering is directed toward Defence, Industrial Automation, Critical Infrastructure, Smart Mobility, Scientific Instrumentation and autonomous cyber-physical systems, where dependable embedded intelligence is fundamental to mission success.

Core technology domains.

01

Sensing and data acquisition

Where physics becomes data. We design sensor interfaces, analog front-ends and signal-conditioning chains that capture vibration, acoustic, biomedical and environmental signals cleanly, at the sampling rate the application demands.

  • Sensor interfaces
  • Signal conditioning
  • High-speed DAQ
02

Signal processing and analytics

Raw streams are reduced before they travel. Spectral estimation, adaptive filtering, compressive sensing and sensor fusion pull out the few features that matter, so less data is stored, sent and powered.

  • Spectral estimation
  • Adaptive filtering
  • Condition monitoring
03

Embedded hardware and firmware

Boards and firmware designed together for predictable timing. Real-time, low-power and built to run unattended for years, from microcontroller nodes to GPU-class edge modules.

  • PCB design
  • Real-time firmware
  • FPGA prototyping
04

Simulation and modelling

Prove it before it is built. We model sensors, signal chains and embedded systems in software, so algorithms, firmware and system behaviour are tested against realistic conditions long before hardware exists.

  • Digital twins
  • Hardware-in-the-loop
  • Model validation
05

Resilient networking

Links that degrade gracefully. Bandwidth-aware protocols over LoRa, Bluetooth and wireless sensor meshes keep fleets of devices talking, and keep each one useful when the link drops.

  • LoRa
  • Bluetooth
  • Wireless sensor networks
06

Edge AI

Models small enough to live on the device. Few-shot learning and model optimization let perception and decisions run within tight memory, power and latency limits, with no cloud round trip.

  • Few-shot learning
  • Model optimization
  • Autonomous perception

Why we build, and where it leads.

Our mission

Mission

To design, build and deploy indigenous embedded intelligence, from the sensor to the decision, so that India's defence, industry and infrastructure run on technology that is made, understood and trusted at home.

Vision

A future where intelligent systems understand, analyse and respond to their environment independently, regardless of connectivity or computing limits, powered by Indian engineering.
The next generation of intelligent systems will not depend on continuous cloud connectivity. They will perform critical sensing, perception, signal processing and decision-making directly on the device, which lowers latency, saves bandwidth, strengthens cybersecurity, reduces power use and improves reliability.

Our vision

Our OngoingResearches.

Acoustic intelligence, aerospace noise control, traffic, industrial testing and wireless sensing: explored in the lab, built for the field.

Select a topic to explore

Compressive sensing

Compressive sensing

Capturing signals with far fewer samples, so systems store, send and power less. By recovering the full signal from a sparse set of measurements, a node can sleep more, transmit less and still see everything that matters.

  • Sparse recovery
  • Fewer samples
  • Lower power
Statistical signal processing and DSP

Statistical signal processing and DSP

Pulling reliable information out of noisy, real-world signals in real time. Estimation and detection methods turn messy sensor streams into clean, trustworthy measurements, running within the timing limits of the device itself.

  • Estimation
  • Detection
  • Real-time DSP
Few-shot learning and model optimization

Few-shot learning and model optimization

Compact models that learn from little data and fit tight memory and power limits. Instead of needing huge labelled datasets, these models adapt from a handful of examples and are then trimmed down to run directly on the edge device.

  • Few-shot learning
  • Model compression
  • On-device inference
Adaptive noise cancellation

Adaptive noise cancellation for non-linear and non-stationary environment

Filters that adapt on the fly to suppress unwanted noise in aerospace and industrial settings. As the noise changes, the filter keeps retuning itself, so the signal of interest stays clear even in loud and unpredictable environments.

  • Adaptive filters
  • Aerospace
  • Industrial

Collaborating across industry, academia and government.

NVIDIA Inception
ANRF consortium
NVIDIA Inception
ANRF consortium
01

Memberships

Programmes and consortia we are part of.

NexmeshIQ Solutions
Enovanto Management
INSPRAXIS
HP Soft Tech
VLSI Lab
02

Engineering partners

We build alongside companies.

TIDES, IIT Roorkee
SWAN Lab, IIT Kharagpur
TIDES, IIT Roorkee
SWAN Lab, IIT Kharagpur
03

Outreach partners

Labs and incubators that extend our reach.

IIT Kharagpur (SWAN Lab)
IIT Bhubaneswar
IIT Dhanbad (ISM)
University of Calcutta
04

Academic and research collaborations

Research with leading institutions.

Working with institutions for societal impact.

NVIDIA Inception Logo NVIDIA Inception
PROGRAMME

NVIDIA Inception

Member of the NVIDIA Inception programme for AI startups.

MeitY Logo MeitY
PROGRAMME

Ministry of Electronics & Information Technology (MeitY)

TIDE 2.0.

SGSITS Logo SGSITS
TIDES Logo TIDES
INCUBATION HOSTS

Incubation hosts

SGSITS Incubation Forum, Indore • TIDES, IIT Roorkee

DST Logo DST
GRANTS

Department of Science & Technology (DST)

iTBI Ignition Grant and NIDHI-EIR.

Our journey.

Drag, scroll or use the arrows. Our little engineer will walk you through it.

APRIL 2025

Company registered

Cognicity Private Limited is registered.

2025

Grants and recognition

Government innovation grants support the BSP platform through prototype, proof of concept and MVP. Founder recognized at TIDE 2.0 2025.

2025 ONWARD

Platforms take shape

EdgeDesk,BSP,Synco and IoTVerseEdgeX take shape.

INDUSTRY AND CONSULTANCY

Industrial projects

Smart Poultry Management, Nex-Brain, Acoustic Emission NDT, Oxygen Purity Module.

RESEARCH

Research programmes

SpectraMed, SGPOR and Weaver progress.

MAY 2026

Patent application filed

Patent application filed for EdgeDesk's autonomous edge-node orchestration.

TODAY

Still building

Continued work across Edge AI, wireless, signal processing and low-power long-range systems.

Cognicity
01 / 07

Meet the team.

Piyush Pandey
LEADERSHIP

Applied research engineer in system architecture design, signal processing and engineering, low power long range communication engineering. He has led the research, design and development of multiple indigenous technologies, from defence technologies to healthcare technologies and AI-driven IoT platforms. Oversees the research facility for RF.

Piyush Pandey

Co-Founder and CEO

Parag Parmar
LEADERSHIP

An engineering professional specializing in advanced embedded systems, high-performance controllers, long-range wireless communication, IoT, industrial automation, and edge AI. Experienced in product design and development, PCB design, hardware-software integration, system testing, and validation, with a focus on building reliable, high-performance indigenous technology solutions for defence, industry, healthcare, and smart infrastructure.

Parag Parmar

Co-Founder and COO

Vivek Vyas
LEADERSHIP

AI engineer specializing in machine learning, deep learning, computer vision, signal processing and intelligent automation. He leads the development of scalable AI across computer vision, predictive analytics and Edge AI, and oversees customer interactions, requirement analysis, stakeholder coordination and end-to-end project management.

Vivek Vyas

Co-Founder and CTO

Bring us the hard part.

A sensor that needs a brain, a prototype that needs a partner, or a research idea that needs a lab. Tell us where you are, and we will map the road from there.

Contact Us ↗ DEFENCE · INDUSTRY · SMART INFRASTRUCTURE · ACADEMIA · GOVERNMENT