Tech for Good
for people, communities and causes.
Explore independent AI agents, open-source tools, skills, plugins and white papers.
Tools and writing with a clear purpose: practical access, learning support, animal welfare and responsible AI.
Indian legal information in plain language
Explains Indian legal processes and helps people identify a sensible starting point. It provides general legal information, not legal advice or representation.
Find the right authority and prepare the next step
Turns a local civic problem into a clear route: the responsible authority, the escalation path and a draft letter or tweet the user can edit and send.
RTI, grievances and public accountability
A multi-agent suite for RTI requests, grievance routes, representative information and election transparency.
For parents and teachers supporting learners aged 13+
A voice-first AI agent that creates short practice plans with clear instructions for learners aged 13 and above, plus practical guidance for parents and teachers.
A practical guide to the ABC process
Explains India's Animal Birth Control process in clear steps for people supporting community animals.
Shared records for community animal care
A community record for identifying street animals, keeping care history together and helping local volunteers coordinate follow-up.
A trust boundary before AI agents act
ATLAS is a small reference implementation that sits between an AI agent and the systems it can change. It checks provenance and instruction-like content, then records an allow, quarantine or block decision.
Use fewer tokens and reduce avoidable AI demand
Shows simple ways to remove repeated prompts and unnecessary context, alongside the energy and cooling resources behind AI use.
Premortem an important decision before taking it
An AI plugin that tests a decision against possible future outcomes, risks and unintended consequences before the user acts.
A few checks before an AI workflow can act
A short self-check for small teams to identify where an AI workflow needs clearer permissions, human review and an action record.
Ten small skills for clearer everyday AI work
Ten public-safe skills for clarifying requests, reducing prompt waste, compressing long outputs, challenging weak ideas, improving writing and handing work over cleanly.
A staged multi-agent workflow with visible review
Separates research, review, revision and final reporting into named stages, so people can see how the work moved and where it was checked.
ATLAS and the semantic trust boundary
A framework for placing a deterministic trust boundary between an AI agent and the systems it can change.
A controlled operating model for telco agents
A framework for designing and governing agentic AI in telecommunications organizations.
BSS for networks where AI agents transact
A framework for the business systems behind networks where AI agents transact.
A 2024 map of practical telco agents
A practical map of AI agents for telecommunications organizations.
A plain-language series for enterprise teams
A plain-language series for enterprise teams working with generative AI.
Identity in an internet shared with agents
Thought leadership on identity in an internet shared with agents.
What cheap output does to professional-services pricing
Thought leadership on what cheap AI output does to professional-services pricing.
A Vedantic comparison of mind and machine
A Vedantic comparison of mind and machine.
Why slow procurement works badly for fast-moving AI
Thought leadership on why slow procurement works badly for fast-moving AI.
TM Forum Inform · Published 13 November 2025
An external publication on the role of MCP in telecommunications.
CIOInfluence · Published 9 April 2025
An external publication on practical agentic AI for telecommunications organizations.
A conversational exploration of the Bhagavad Gita
What if the questions you face today could be explored through Krishna's teachings? Gita Tutor offers a conversational way to study and reflect on the Bhagavad Gita.
Items without supplied URLs remain non-clickable. Metrics are editorial reach labels, not product-user counts.