Guests have been shocked, audio system have been exhilarating, and the AI neighborhood was extra related than ever earlier than, simply as India’s most futuristic occasion ended on Friday, August 22, 2025. It was to create a course for the way forward for AI in India and past. From voice-driven interactions to no-code workflows. From defending AI techniques to scaling multi-agent collaboration. Within the closing chapter of the summit, members have been impressed, knowledgeable and keen to construct.
For those who missed the early days of DHS 2025, you possibly can learn all in regards to the occasions of DHS 2025 Day 1 and Day 2 right here.
DHS 2025, Day 3: When genai meets precise impression
The day started with Vijay Gabare, co-founder and CPO of Infilect. He gave an inspiring session on why Gunai and LLMS fail and the way tweaks might help. In a related real-world instance, he reminded the viewers that even probably the most refined fashions come across primary logic. Nevertheless, light-weight, domain-specific tweaks permit these similar techniques to be transformed into precision instruments. His message was clear. The AI would not shine nicely sufficient. It additionally must be dependable.
Persevering with momentum from DHS 2025, Day 3 of AG2 co-creator and co-founder Chi Wang, DHS 2025, scaled AI by means of Massgen: Multi-Agent Collaboration. His reside demos – a number of AI brokers fixing issues – was a vivid showcase of how collaboration in addition to scale defines the subsequent frontier of AI. For lots of the viewers, this was a glimpse into how multi-agent techniques may mirror human teamwork and converge right into a superior resolution by means of collective reasoning.
India’s Sovereignty AI Imaginative and prescient
Pratyush Kumar, co-founder of Sarvam AI, some of the anticipated voices of the day, gave his keynote tackle and created the Indian AI ecosystem. From imaginative and prescient to sovereignty. His name for motion was profound. India should not be remoted, however construct with strategic autonomy. It will hold computing, information and innovation rooted within the nation. His passionate enchantment to expertise improvement and home funding made him painting the longer term wherein India is just not a participant, however a shaper within the international AI story.
Remove complexity and improve innovation
No periods gained the spirit of accessibility than Mayank Aggarwal, co-founder and CEO of Evolvue AI. With humor, sensible demos and real-world case research, he confirmed that no-codes are usually not “no energy.” As an alternative, it’s a revolution that places AI improvement into everybody’s palms, breaks limitations and democratizes innovation on a big scale.
Audio: New command
Subsequent was the session that felt like he was peering into the way forward for human-computer interplay. Microsoft’s Ranjani Mani, Director and Nation Head – Generative AI and AI Architect’s Manoranjan Rajguru lit up the stage with the interplay of AI Voice Brokers: The Way forward for Human-Laptop. From customer support to healthcare, they demonstrated how voice-first interface redefines communication with machines. From clicks to conversations, make expertise extra human than ever.
Constructing belief by means of safety and governance
The dialog then shifted to accountability and resilience. Satnam Singh and Shivaraj Mulimani, safety information scientists at Acalvio Applied sciences, have introduced their safety system from the groundbreaking session Pink Teaming Genai: Inside Out. They confirmed that predicting the menace earlier than they assault was the one step ahead. The session unpacked a technique to probe the Genai system for vulnerabilities and strengthen them in opposition to enemies. Their message resonated deeply: in an innovating race, safety can’t be an afterthought.
Smarter cities, smarter techniques
From safety, the highlight has moved to infrastructure. Manpreet Singh, a senior information scientist at Microsoft, electrified his viewers with the creation of real-time multi-agent AI for public journey techniques. His demonstrations from Rahat, an AI system that helps passengers on railways and airports, gave a glimpse right into a seamless, clever, and accessible future with AI-powered public journey techniques. It was an innovation designed for on a regular basis commuters, not for elites.
Framework battle
Then got here the second when many individuals have been ready: Autogen vs Crewai vs Langgraph: Battle of the Agent Frameworks. Mayank Aggarwal, Praneeth Paikray and Sanathraj Narayan ignited an emotional showdown and attacked one another three agent frameworks. Every framework launched distinctive strengths, and the vigorous dialogue made members vigorous with new concepts. This was a dialogue that prompted the neighborhood to rethink how agent AI is constructed, optimized, and expanded.
When an agent speaks the identical language
Avinash Pathak, senior AI engineer at Nvidia, has moved the baton ahead with the agent protocol with the agent: advantages and workflow. His speech was a imaginative and prescient for a future the place AI brokers may talk and cooperate seamlessly between techniques. He left a robust level for his viewers by exhibiting how interoperability locks effectivity. The longer term is about smarter ecosystems.
Trade on the intersection of AI and impression
The business use circumstances on the third day have been simply as thrilling because the framework. In Session Multimodal Genius for Vitality Infrastructure Inspection Report, Shell’s main AI lead, Prakalp Somawanshi demonstrates mix imaginative and prescient and language fashions to remodel vitality sector inspections to boost security, effectivity and compliance.
Emirates Senior Software program Engineer Pradeep Kumar has introduced from Imaginative and prescient to Motion: Multimodal Agent AI in Precise Use. His session proved that this isn’t about futuristic desires, however that AI techniques have already got a concrete impression. From aviation to logistics and retail.
Creativity, collaboration, and so forth.
The ultimate session led the viewers to the centre of creativity and collaboration. Adobe machine studying engineer Dhruv Nair explored open instruments for Inventive AI Brokers: Collaborative media creation. He confirmed how AI will develop into a co-pilot for designers, storytellers and creators. In the meantime, Luis Serrano, founder and chief schooling officer of Serrano Academy, captivated members with a visible information to the eye mechanisms of LLMS. This session reworked advanced ideas into elegant, visible narratives.
Collectively, these periods reminded the neighborhood that AI isn’t just about effectivity. It is about enhancing creativeness and human potential.
AV Luminary Awards: Honoring the pioneer on the third day
The day closed with celebration and applause because the AV Laminal Award acknowledged contributors from the Prime 7 AI neighborhood. These leaders have pushed the AI neighborhood with dedication, collaboration and impression.
Dr. Joshua Starma
Bhavishya Pandit
Dipanjang S.
Lakshmi Devi Prakash
Abhishek Divekar
Sarvagya Agrawal
Avinash P.
Their notion was a tribute to the spirit of collective progress that Datahack Summit embodies.
DHS 2025: Shut Reflection
Day 3 of DHS 2025 was a becoming finale on your imaginative and prescient, vitality and pace over the course of three days. From Indian sovereign AI to framework battles, from securing techniques to growing improvement, the dialog that came about at Knowledge Hack Summit 2025 echoes all the way in which after the lights dim.
As a result of DHS isn’t just what AI can do right this moment. It is about what we, as a neighborhood, make doable tomorrow.
Log in and proceed studying and revel in skilled curated content material.
Proceed studying at no cost
