Selected work / 2022—26

Built for the moment after launch.

Engagements across product, AI, infrastructure, and automation. The common thread: useful software that keeps getting better when real people use it.

Selected work

Selected systems, shaped around the people using them.

A search layer for the company's scattered memory project preview

A search layer for the company's scattered memory

Anvaya is an AI-powered enterprise assistant that provides natural-language access to organisational data, knowledge, and connected tools. It can handle multi-step analytical queries, retrieve information across multiple sources, execute authorised operations, and adapt its capabilities to each user's role and permissions. Built as a multi-tenant PWA with configurable AI providers, caching, and controlled data retention, Anvaya turns complex enterprise information workflows into a single conversational interface.

PythonRAGPostgreSQL
View case study
Making government schemes easier to discover, understand, and access project preview

Making government schemes easier to discover, understand, and access

Yojana Sahayak is a multilingual platform that makes government welfare schemes easier to discover, understand, and access. We built the full-stack platform, an automated government-scheme data collection pipeline, and an AI assistant supporting text and voice interactions across eight languages.

GovTechReactAIWeb ScrapingMultilingual
View case study
Fast, reliable notification delivery across every channel project preview

Fast, reliable notification delivery across every channel

Sandesh is a lightweight multi-channel notification delivery system built to reliably send push notifications, email, and SMS while supporting priority-based queuing, retries, scheduling, templating, and failure handling.

NotificationsBackendQueuesReliabilitySystems
View case study
Computer vision for safer workplaces project preview

Computer vision for safer workplaces

Sentinel is a real-time computer-vision system for automated PPE compliance monitoring. It detects workers, helmets, and safety vests from images and camera feeds and identifies missing protective equipment as potential safety violations.

Computer VisionYOLOPPE DetectionAISafety
View case study
High-throughput inference infrastructure for unpredictable traffic project preview

High-throughput inference infrastructure for unpredictable traffic

Vega is a high-throughput inference serving system designed for extreme traffic bursts on fixed private infrastructure. Instead of relying on reactive autoscaling, it uses shared queuing, dynamic batching, backpressure, and controlled load shedding to maximise existing compute utilisation.

AI InfrastructureInferenceHigh ThroughputPythonDistributed Systems
View case study
Continuous competitive intelligence for automotive inventory project preview

Continuous competitive intelligence for automotive inventory

Drishti is a competitive intelligence platform built for a US used-car dealership. It continuously monitors competitor inventory, performs detailed vehicle-level comparisons, analyses pricing and listing quality, profiles competing dealerships, and turns changing market data into actionable recommendations.

Competitive IntelligenceWeb ScrapingAutomotiveMarket AnalysisAI
View case study
Private semantic knowledge retrieval, fully offline project preview

Private semantic knowledge retrieval, fully offline

Bodh is an offline semantic knowledge retrieval system that lets organisations build private, customisable knowledge bases from technical documentation, internal practices, tool information, and company-specific knowledge. It indexes knowledge in a learned latent space and retrieves information based on meaning rather than exact keyword matches.

Semantic SearchInformation RetrievalOffline AIKnowledge SystemsMachine Learning
View case study
An AI research assistant for finding, understanding, and connecting academic knowledge project preview

An AI research assistant for finding, understanding, and connecting academic knowledge

Shodh is an AI-powered research assistant built around retrieval-augmented generation. It helps researchers explore academic literature through natural-language questions, retrieve relevant research context, and generate grounded responses from the available knowledge base.

RAGResearch AssistantNLPSemantic SearchAI
View case study
An end-to-end customer intelligence platform project preview

An end-to-end customer intelligence platform

SmartGrowthAI is an end-to-end customer intelligence platform connecting data foundations, machine-learning models, API serving, and business dashboards. It combines churn prediction, customer risk analysis, demand forecasting capabilities, and NLP-based insights from customer complaints.

Customer IntelligenceMachine LearningChurn PredictionForecastingNLP
View case study
Intelligent product label extraction from packaging images project preview

Intelligent product label extraction from packaging images

Visionary Extractor is a computer-vision-based product label extraction system that turns packaging images into structured, machine-readable product information. It provides a foundation for product cataloguing, inventory workflows, packaging verification, and supply-chain intelligence.

Computer VisionOCRProduct IntelligenceSupply ChainPackaging
View case study
Explainable Hotel Review Intelligence project preview

Explainable Hotel Review Intelligence

Atithi is a full-stack machine-learning platform for real-time hotel review sentiment analysis. It combines NLP preprocessing, a Voting Classifier ensemble, aspect-based sentiment analysis, and LIME explainability to turn guest reviews into interpretable customer-experience insights.

NLPSentiment AnalysisExplainable AIHospitalityMachine Learning
View case study
11projects shipped
10domains explored
01small senior team

Building something
difficult?

Tell us what you're building, improving, or trying to figure out.

Start a project See our work