Rohit Joshi

Region Head at Fractal Analytics, Australia and NZ

[Fractal @fractalai]

What actually makes enterprise AI intelligent enough to reason, retrieve, and respond with context?

It starts with pre-training.

While RAG, fine-tuning, and AI agents often get the spotlight, pre-training is the foundation that enables models to understand language, learn patterns, capture context, and support downstream enterprise AI applications at scale.

In this article, Sandeep Gore breaks down why pre-training is critical to building reliable enterprise AI systems, from semantic search and conversational AI to RAG and intelligent agents.

Read the full article to understand why pre-training is not just a technical step but the foundation of enterprise AI.

https://t.co/GbmpvS7WtY

#EnterpriseAI #GenerativeAI #PreTraining #FoundationModels #RAG #AITransformation

[Fractal @fractalai]

As the world's demand for AI intensifies, the spotlight is shifting to where it matters most: India's role as the next global AI destination.

Our Group CEO and nasscom Chairperson, Srikanth Velamakanni, has been front and center in these conversations, from data centers and tech sovereignty to geopolitical shifts reshaping the digital economy.

Here's where you can catch him shaping the narrative:

🔸𝗡𝗗𝗧𝗩 𝗣𝗿𝗼𝗳𝗶𝘁: 𝗔𝗜 𝗪𝗶𝗹𝗹 𝗠𝘂𝗹𝘁𝗶𝗽𝗹𝘆 𝗜𝗻𝗱𝗶𝗮𝗻 𝗘𝗰𝗼𝗻𝗼𝗺𝘆
How global AI companies are scouting India for datacenters, and why falling software costs mean surging demand for Indian talent and innovation.

https://t.co/5dMdkL3kwM

🔸𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗧𝗼𝗱𝗮𝘆: 𝗧𝗵𝗲 𝗖𝗮𝘀𝗲 𝗳𝗼𝗿 𝗧𝗲𝗰𝗵 𝗦𝗼𝘃𝗲𝗿𝗲𝗶𝗴𝗻𝘁𝘆
A deep dive into India's tech stack, dependencies, and the path forward.

"The world is less globalized today than it was in 2015. Resilience and technological sovereignty are far more important than they once seemed."

https://t.co/TSVgHkaSLk

🔸𝗜𝗻𝗱𝗶𝗮 𝗧𝗼𝗱𝗮𝘆 𝗙𝗶𝗻𝗮𝗻𝗰𝗶𝗮𝗹 𝗦𝘂𝗺𝗺𝗶𝘁 𝟮𝟬𝟮𝟲
Leading conversations on tech, policy, and enterprise transformation.

https://t.co/twrwf3vK6J

🔸𝗡𝗗𝗧𝗩 𝗣𝗿𝗼𝗳𝗶𝘁: 𝗚𝗲𝗼𝗽𝗼𝗹𝗶𝘁𝗶𝗰𝘀 & 𝗜𝗻𝗻𝗼𝘃𝗮𝘁𝗶𝗼𝗻
Geopolitical shifts, market risks, and what's next for tech. SpaceX, India's economy, and the forces reshaping global business.

https://t.co/twrwf3vK6J

The conversation is evolving. The stakes are rising. And India's moment is now.

#EnterpriseAI #AI #GenAI #DecisionIntelligence #AITransformation #FractalinNews

[Fractal @fractalai]

The demo isn't where enterprise AI deals die. It's the six weeks between "we love this" and a working pilot in the carrier's cloud. That gap, IT tickets, access management, firewall configurations, and vendor provisioning, is where deals go cold. Not because anyone stopped caring. Because friction accumulates faster than momentum.​

𝗖𝗼𝗴𝗲𝗻𝘁𝗶𝗾 𝗨𝗻𝗱𝗲𝗿𝘄𝗿𝗶𝘁𝗶𝗻𝗴 deploys differently.​

One IT person. A GitHub repo. A one-page runbook. The full stack: Data Triage, Risk Analyzer, UW Copilot, QA Agent, stands up inside the carrier's own Azure tenant in a few hours. No vendor access required. No dependency on our team.​

"Can we try this?" is now a same-day answer. For CTOs and CIOs evaluating AI in the underwriting stack: how long does it take to go from demo to real data in your environment? That evaluation on real data matters more than any benchmark on a slide.​

Time-to-pilot is the hardest variable to move in enterprise insurance software. We moved it from weeks to hours.​

Swipe for how it works →

#UnderwritingAI #EnterpriseAI #InsuranceInnovation #Cogentiq

[Fractal @fractalai]

Walk into any supermarket today and you’ll notice how every product is trying to feel relevant to you.

CPG has evolved from mass messaging to segmentation to real-time conversations. But today, even that isn’t enough. Consumers expect brands to anticipate their needs and respond instantly.

AI is making that shift possible by turning data into real-time intelligence, enabling brands to move from reaction to prediction.

What’s emerging is not a replacement of humans, but a powerful partnership.

Because now, every product has to speak to everyone.

Read the feature story in aisight Vol. 11: https://t.co/zzyBPJ07eq

#AI #CPG #GenAI #ConsumerInsights #DigitalTransformation #aisight

[Rohit Joshi @rojoshi75]

@Alokkumarzz This is good lecture, I shared this with my teenage son.

[Rohit Joshi @rojoshi75]

@Gosleepriya 18

[Rohit Joshi @rojoshi75]

@samrac92 2,3,1

[Fractal @fractalai]

We just closed Q3 FY26, strong quarter as a newly listed company.

🔸 ₹854.4 Cr revenue (+21% YoY)
🔸 ₹152.1 Cr Adjusted EBITDA (+24% YoY); 17.8% margin
🔸 ₹100.1 Cr Profit After Tax
🔸 47.2% Gross Margin, best-in-class
🔸 30% CAGR over the past decade
🔸 Net Revenue Retention at 114%; NPS at 77

Growth was led by strong momentum in Healthcare & Life Sciences (+78% YoY) and Banking & Financial Services (+26% YoY); a clear proof that enterprise AI adoption is accelerating across industries.
Our client partnerships continue to deepen:

🔸 6 clients generating $20M+ annually (up 2 YoY)
🔸 58 clients generating $1M+ annually (up 8 YoY)

On the innovation front, our products are setting global benchmarks:
https://t.co/RD8ysSvrz2 2.0 became the first AI model to score 50+ on OpenAI’s HealthBench (Hard), outperforming leading foundation models.

PiEvolve ranked among the top-performing agents on OpenAI’s MLE-Bench, as of 24 February 2026, putting it ahead of agents from major global AI labs.
This quarter reflects what we’ve been building toward for years:
🔸 Deep client trust
🔸 Consistent, profitable growth
🔸 A disciplined, high-performance enterprise AI company

As we scale post-IPO, our focus remains unchanged - powering every human decision in the enterprise and creating sustained long-term value for our clients and shareholders.

Grateful to our clients, partners, and 5,000+ Fractalites worldwide who make this possible. We’re just getting started.

Learn more in our earnings release:

𝐈𝐧𝐯𝐞𝐬𝐭𝐨𝐫 𝐑𝐞𝐥𝐚𝐭𝐢𝐨𝐧𝐬: https://t.co/eaE7WAkYHD

𝐈𝐧𝐯𝐞𝐬𝐭𝐨𝐫 𝐏𝐫𝐞𝐬𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧: https://t.co/h7bGBUIGbK

𝐋𝐞𝐭𝐭𝐞𝐫 𝐭𝐨 𝐒𝐡𝐚𝐫𝐞𝐡𝐨𝐥𝐝𝐞𝐫𝐬: https://t.co/ihYLX5piBc

𝐅𝐚𝐜𝐭 𝐒𝐡𝐞𝐞𝐭: https://t.co/35soUoKGDY

𝐏𝐫𝐞𝐬𝐬 𝐑𝐞𝐥𝐞𝐚𝐬𝐞: https://t.co/SVwU9tLvsJ

#Fractal #EnterpriseAI

[Fractal @fractalai]

𝐒𝐩𝐞𝐜𝐢𝐚𝐥 𝐒𝐞𝐬𝐬𝐢𝐨𝐧: 𝐀𝐠𝐞𝐧𝐭𝐢𝐜 𝐀𝐈 𝐏𝐥𝐚𝐲

NASSCOM Technology & Leadership Forum (NTLF) – Day 1
Moderated by Shouvik Das (Mint)
Speakers: Babak Hodjat, Kishor Patil, Hari Balaji, Srikanth Velamakanni

If the first session at NTLF set the context for exponential change, the second session zoomed in on a sharper question:

👉 Is Agentic AI actually delivering business value, or are we still in pilot purgatory?
The discussion was refreshingly honest. No hype. No buzzword overuse. Just a grounded exploration of what it really takes to move from experimentation to enterprise-scale impact.

Here are the key reflections:

1️⃣ 𝐀𝐈 𝐢𝐬 𝐚𝐛𝐨𝐮𝐭 𝐝𝐞𝐜𝐢𝐬𝐢𝐨𝐧 𝐢𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞 𝐚𝐭 𝐬𝐜𝐚𝐥𝐞
One of the most powerful reframes from the panel:
AI is not just automation.
It’s about building a 𝐥𝐚𝐲𝐞𝐫 𝐨𝐟 𝐢𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞 𝐭𝐡𝐚𝐭 𝐩𝐨𝐰𝐞𝐫𝐬 𝐦𝐢𝐥𝐥𝐢𝐨𝐧𝐬 𝐨𝐟 𝐝𝐞𝐜𝐢𝐬𝐢𝐨𝐧𝐬 𝐞𝐯𝐞𝐫𝐲 𝐝𝐚𝐲.
If you can improve the quality, speed, or consistency of decisions, even marginally, the impact compounds across the enterprise.
And as AI accuracy improves, the number of processes where AI can match or exceed human performance keeps expanding.
That changes the game.

2️⃣ 𝐄𝐯𝐞𝐫𝐲 𝐞𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐩𝐫𝐨𝐜𝐞𝐬𝐬 𝐢𝐬 𝐧𝐨𝐰 “𝐎𝐩𝐞𝐧” 𝐭𝐨 𝐫𝐞𝐢𝐧𝐯𝐞𝐧𝐭𝐢𝐨𝐧

A big theme that emerged:
As AI becomes more capable, the total addressable opportunity expands.

🔸Business processes
🔸Products
🔸Services
🔸Legacy tech stacks

Everything is now a candidate for reinvention.
But this isn’t just about layering AI on top of broken systems. It’s about:

🔸Reimagining workflows
🔸Rebuilding clunky enterprise tech
🔸Creating an intelligence layer across data

And that requires engineering discipline, not pixie dust.

3️⃣ 𝐀𝐠𝐞𝐧𝐭𝐢𝐜 𝐀𝐈 ≠ 𝐌𝐚𝐠𝐢𝐜. 𝐈𝐭’𝐬 𝐞𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐞𝐝 𝐚𝐮𝐭𝐨𝐧𝐨𝐦𝐲.

One of the clearest explanations of Agentic AI during the session:
Agentic AI gives systems 𝐚𝐠𝐞𝐧𝐜𝐲;
The ability to reason, plan, and act, not just generate.
But with that autonomy comes responsibility.
You can’t just “spread LLM dust” across your enterprise and expect transformation.
Agentification is:

🔸Designed
🔸Structured
🔸Measured
🔸Iteratively engineered

The science of building scalable, reliable multi-agent systems is still evolving. And organizations must treat it as a serious engineering discipline.

4️⃣ 𝐓𝐡𝐞 𝐑𝐎𝐈 𝐝𝐞𝐛𝐚𝐭𝐞 𝐢𝐬 𝐫𝐞𝐚𝐥, 𝐚𝐧𝐝 𝐧𝐮𝐚𝐧𝐜𝐞𝐝

There was a healthy tension in the room around ROI.
On one hand:
🔸Enterprises want clear, measurable returns.
🔸Financial strain makes caution understandable.
🔸In regulated industries like mobility, deterministic outcomes are non-negotiable.

On the other hand:
🔸AI capabilities are evolving so quickly that yesterday’s failed pilot might succeed today.
🔸Many companies are underinvested but impatient.
🔸We may be asking ROI questions too early in some cases.

One interesting perspective:
At an individual company level, ROI matters immediately.
At an economic or national level, AI is infrastructure, like railways or highways. You build first, returns compound over time.
That distinction was powerful.

5️⃣ 𝐍𝐨𝐭 𝐫𝐯𝐞𝐫𝐲𝐭𝐡𝐢𝐧𝐠 𝐧𝐞𝐞𝐝𝐬 𝐀𝐠𝐞𝐧𝐭𝐢𝐜 𝐀𝐈
𝐀 𝐜𝐫𝐢𝐭𝐢𝐜𝐚𝐥 𝐫𝐞𝐚𝐥𝐢𝐭𝐲 𝐜𝐡𝐞𝐜𝐤:
There’s confusion between:
🔸Generative AI
🔸Agentic AI
🔸Machine learning
🔸Rule-based systems
Some processes are deterministic and may not need over-engineered autonomy.

But here’s the nuance:
Agentic systems can incorporate rules. They’re not mutually exclusive.

The real skill lies in deciding:

🔸What belongs in code
🔸What belongs in models
🔸What belongs in governance

That architectural judgment will differentiate winners.

6️⃣ 𝐓𝐡𝐞 𝐡𝐚𝐫𝐝𝐞𝐬𝐭 𝐩𝐫𝐨𝐛𝐥𝐞𝐦 𝐢𝐬 𝐜𝐮𝐥𝐭𝐮𝐫𝐞, 𝐧𝐨𝐭 𝐭𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐲
Data readiness.
Process redesign.
Infrastructure modernization.
All important.
But the deeper challenge?

Building an𝐀𝐈-𝐟𝐢𝐫𝐬𝐭 𝐜𝐮𝐥𝐭𝐮𝐫𝐞.
Everyone has used AI personally.
That creates an illusion of understanding.

But enterprise AI transformation is top-down, strategic, and systemic.
It requires coherent internal communication, aligned KPIs, and leadership conviction.

Without that, pilots remain pilots.

7️⃣ 𝐓𝐡𝐞 𝐬𝐡𝐢𝐟𝐭 𝐟𝐨𝐫 𝐬𝐞𝐫𝐯𝐢𝐜𝐞𝐬 𝐜𝐨𝐦𝐩𝐚𝐧𝐢𝐞𝐬

Will agentic AI reduce the role of services firms?
The answer from the panel was clear:
Yes, disruption is coming.
But so is opportunity.
The knowledge of domain workflows, enterprise complexity and last-mile integration remains critical.
What changes:
🔸Project cycles must be more agile.
🔸Five-year waterfall programs don’t survive in an AI world.
🔸Engineering shifts from deterministic software modules to semi-autonomous intelligent Lego blocks.
But assembling those blocks into reliable, measurable business systems?
That still requires deep expertise.

8️⃣𝐈𝐧𝐝𝐮𝐬𝐭𝐫𝐲-𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐫𝐞𝐚𝐥𝐢𝐭𝐢𝐞𝐬 𝐦𝐚𝐭𝐭𝐞𝐫

In mobility and regulated sectors:
🔸Validation infrastructure is not AI-ready.
🔸Architectures were not designed for AI ecosystems.
🔸Financial caution slows experimentation.
AI-first design in new programs will unlock far more value than retrofitting legacy systems.
That long-term shift will be substantial.

𝐅𝐢𝐧𝐚𝐥 𝐑𝐞𝐟𝐥𝐞𝐜𝐭𝐢𝐨𝐧
Summarizing this session:

Agentic AI is not about replacing processes.
It’s about redesigning intelligence inside them.
The disruption will be significant,
But only for organizations that engineer it deliberately.
We are still early.
The science of large-scale agentification is forming.
The cultural shift is incomplete.
But the direction is unmistakable.

NTLF Day 1 made one thing clear:
The future of enterprise is not just AI-enabled.
It is multi-agent, engineered, and continuously evolving.

The question now isn’t whether to adopt Agentic AI.
It’s how thoughtfully and how fast we choose to build with it.

#NTLF2026 #NTLF26 #IndiasEdge #TechDrivenHumanCentered
#IndiasEdgeatNTLF2026

@srikanth

[Fractal @fractalai]

𝐄𝐯𝐞𝐫𝐲𝐭𝐡𝐢𝐧𝐠, 𝐄𝐯𝐞𝐫𝐲𝐰𝐡𝐞𝐫𝐞, 𝐀𝐥𝐥 𝐚𝐭 𝐎𝐧𝐜𝐞: 𝐋𝐞𝐚𝐝𝐢𝐧𝐠 𝐢𝐧 𝐚 𝐖𝐨𝐫𝐥𝐝 𝐨𝐟 𝐄𝐱𝐩𝐨𝐧𝐞𝐧𝐭𝐢𝐚𝐥 𝐂𝐡𝐚𝐧𝐠𝐞

@nasscom Technology & Leadership Forum (NTLF) – Day 1

Day 1 at NASSCOM NTLF set the tone exactly where it needed to: bold, honest, and forward-looking.
In a powerful opening session featuring Rajesh Nambiar, @srikanth and Sindhu Gangadharan, the conversation centered on a reality we’re all experiencing:

We are not navigating a single disruption.
We are navigating simultaneous acceleration, across AI, geopolitics, enterprise tech, talent models and global economics.

And it’s all happening at once.

Here are the biggest takeaways from the discussion:

1️⃣ 𝟐𝟎𝟐𝟓 𝐰𝐚𝐬 𝐧𝐨𝐭 𝐚 𝐲𝐞𝐚𝐫 𝐨𝐟 𝐫𝐞𝐜𝐚𝐥𝐢𝐛𝐫𝐚𝐭𝐢𝐨𝐧. 𝐈𝐭 𝐰𝐚𝐬 𝐚 𝐲𝐞𝐚𝐫 𝐨𝐟 𝐚𝐜𝐜𝐞𝐥𝐞𝐫𝐚𝐭𝐢𝐨𝐧.

We expected AI adoption to grow. What caught everyone off guard was the speed. Models are evolving every few weeks. Developer roles are shifting from writing code to orchestrating intelligence. Boardroom conversations now include resilience, trust, and geopolitical risk alongside innovation.
The pace is the story.

2️⃣ 𝐀𝐈 𝐡𝐚𝐬 𝐦𝐨𝐯𝐞𝐝 𝐛𝐞𝐲𝐨𝐧𝐝 𝐭𝐡𝐞 𝐡𝐲𝐩𝐞 𝐜𝐲𝐜𝐥𝐞.

Not long ago, we were debating whether scaling laws would hold. Today, we’re discussing recursive AI self-improvement and enterprises moving toward AI-generated code at scale.
In one year, the narrative has shifted from experimentation to structural transformation.

3️⃣ 𝐀𝐈 𝐜𝐨𝐦𝐩𝐫𝐞𝐬𝐬𝐞𝐬 𝐜𝐨𝐬𝐭, 𝐛𝐮𝐭 𝐞𝐱𝐩𝐚𝐧𝐝𝐬 𝐨𝐩𝐩𝐨𝐫𝐭𝐮𝐧𝐢𝐭𝐲.

Yes, modernization costs will drop. Legacy transformation that once cost billions may cost a fraction with AI assistance.

But that doesn’t shrink the market. It unlocks it.
Projects that were too risky, too expensive or too complex are now viable.

First-order thinking says AI reduces revenue pools.
Second-order thinking says AI expands the total addressable opportunity.

That reframing is critical for our industry.

4️⃣ 𝐒𝐚𝐚𝐒 𝐢𝐬𝐧’𝐭 𝐝𝐲𝐢𝐧𝐠. 𝐈𝐭’𝐬 𝐛𝐞𝐜𝐨𝐦𝐢𝐧𝐠 𝐚𝐠𝐞𝐧𝐭𝐢𝐜.

We won’t navigate clunky dashboards forever. We’ll converse with systems.
Ask questions. Trigger workflows. Orchestrate outcomes.

But underneath that conversational layer, enterprise architecture, data integrity and orchestration become even more important, not less.

5️⃣ 𝐄𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐀𝐈 𝐭𝐫𝐚𝐧𝐬𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧 𝐡𝐚𝐬 𝐭𝐡𝐫𝐞𝐞 𝐛𝐢𝐠 𝐩𝐥𝐚𝐲𝐬:

🔸Reimagining business processes end-to-end
🔸Building an intelligence layer across enterprise data
🔸Reimagining the workforce in an AI-first world

And perhaps the most provocative insight from the stage:

𝐄𝐱𝐩𝐞𝐫𝐢𝐞𝐧𝐜𝐞 𝐚𝐥𝐨𝐧𝐞 𝐢𝐬 𝐧𝐨 𝐥𝐨𝐧𝐠𝐞𝐫 𝐭𝐡𝐞 𝐦𝐨𝐚𝐭.
AI-native thinking, extreme talent and fresh perspectives may outpace traditional models of expertise.
For the NASSCOM ecosystem, especially with the “transform the old, accelerate the new” lens — this moment is less about survival and more about reinvention.

There is massive technical debt across industries.
There is enormous pent-up transformation demand.
And there may never be a better moment to address both.

The session reminded us that this is not a shrinking era for technology services or product innovation. It is an expansion era, if we choose to lean into it with clarity, courage, and responsibility.
Everything is changing.
Everywhere.
All at once.

The real question for all of us at NTLF and beyond is:
How do we lead when acceleration itself becomes the constant?

[Fractal @fractalai]

The pace of change has never been linear. But leadership, today, is no longer about speed but sensemaking.

At the India AI Impact Summit 2026, the conversations won’t just stay on stage, they’ll come alive across the pavilions.

Here’s what to watch out for from Fractal:

🔸 At the 𝐁𝐡𝐚𝐫𝐚𝐭 𝐏𝐚𝐯𝐢𝐥𝐢𝐨𝐧, 𝐇𝐚𝐥𝐥 𝟏𝟒, catch AI creating tangible healthcare impact through https://t.co/PbwZHynSir, enabling rapid TB detection in radiology even in remote locations, and
🔸 At 𝐇𝐚𝐥𝐥 𝟏𝟒, 𝐏𝐨𝐝 𝐍𝐨. 𝟑𝟎, explore https://t.co/RD8ysSvrz2 𝐚𝐧𝐝 𝐅𝐚𝐭𝐡𝐨𝐦, India’s Large Reasoning Models, built to make advanced intelligence accessible to all under the IndiaAI Mission.
🔸 At 𝐇𝐚𝐥𝐥 𝟏𝟒, 𝐏𝐨𝐝 𝐍𝐨. 𝟏𝟔, technology meets scale and application with 𝐂𝐨𝐠𝐞𝐧𝐭𝐢𝐪 𝐂𝐗, our agentiq-powered solution reimagining customer experience, and 𝐈𝐪𝐢𝐠𝐚𝐢, an agentiq product designed to assess skills objectively and holistically at scale; one we actively use within our own hiring processes.

If you’re thinking about where India is headed and who is shaping that journey, this is where the conversation begins.

17th - 20th February 2026
📍 Bharat Mandapam, New Delhi

See you there!

#IndiaAIImpactSummit #AIForIndia #FutureOfLeadership

[Image description : Fractal poster announcing participation at India AI Impact Summit 2026 with event dates and venue in New Delhi.]

[Fractal @fractalai]

Join Sandeep Dutta today, as he participates in the panel discussion at the India AI Impact Summit 2026, where the session on 𝐀𝐈 𝐟𝐨𝐫 𝐀𝐥𝐥 – 𝐓𝐡𝐞 𝐁𝐮𝐬𝐢𝐧𝐞𝐬𝐬-𝐥𝐞𝐝 𝐈𝐧𝐜𝐥𝐮𝐬𝐢𝐨𝐧 𝐄𝐧𝐠𝐢𝐧𝐞, will examine how AI-driven business models can be designed to uplift communities, expand access to essential services, and create equitable opportunities, while remaining commercially viable

The discussion will cover:
🔸 𝐒𝐨𝐜𝐢𝐚𝐥 𝐄𝐦𝐩𝐨𝐰𝐞𝐫𝐦𝐞𝐧𝐭 𝐭𝐡𝐫𝐨𝐮𝐠𝐡 𝐌𝐚𝐫𝐤𝐞𝐭 𝐏𝐮𝐥𝐥: How business-driven AI (e.g., credit scoring for the unbanked or AI-led crop advisory for farmers) creates a sustainable loop of social impact.
🔸 𝐅𝐫𝐨𝐦 𝐏𝐢𝐥𝐨𝐭𝐬 𝐭𝐨 𝐏𝐫𝐨𝐟𝐢𝐭: Moving the conversation from "AI hype" to "AI ROI" for small businesses, focusing on efficiency, cost reduction, and market expansion.
🔸 𝐃𝐏𝐈 𝐚𝐬 𝐭𝐡𝐞 𝐆𝐫𝐞𝐚𝐭 𝐄𝐪𝐮𝐚𝐥𝐢𝐳𝐞𝐫: Discuss how India Stack (Aadhaar, UPI, ONDC) can serve as the data backbone for AI. For instance, how ONDC democratizes data access for SMEs to compete with e-commerce giants.
🔸 The Cloud as a Utility: Treating cloud and compute resources not just as IT costs, but as essential infrastructure like electricity, enabling startups to scale without heavy upfront Capex.

If you’re wondering how AI can be both commercially grounded and socially transformative, this is a conversation worth being part of.

𝐈𝐧𝐝𝐢𝐚 𝐀𝐈 𝐈𝐦𝐩𝐚𝐜𝐭 𝐒𝐮𝐦𝐦𝐢𝐭 𝟐𝟎𝟐𝟔

🔸 February 17, 2026
🔸 3:30 PM – 4:25 PM
🔸 Meeting Room 8, Bharat Mandapam, New Delhi

See you there!

#AIImpactSummit2026 #AIForAll

[Image description : Poster announcing Sandeep Dutta speaking at the AI Impact Summit in New Delhi.]

[Rohit Joshi @rojoshi75]

@Qudsia_Malik33 18

[Rohit Joshi @rojoshi75]

@Arden_2210 29

[Rohit Joshi @rojoshi75]

@Arden_2210 18

[Rohit Joshi @rojoshi75]

@Jindo_1307979 63, previous number multiplied by 2 plus 1

[Rohit Joshi @rojoshi75]

Dhanyawad @airindia lovely check in experience @BLRAirport with smiley agent @dhanujkumar27825

[Fractal @fractalai]

𝐃𝐚𝐭𝐚 𝐬𝐜𝐢𝐞𝐧𝐜𝐞 𝐭𝐞𝐚𝐦𝐬 𝐚𝐫𝐞𝐧'𝐭 𝐬𝐡𝐨𝐫𝐭 𝐨𝐧 𝐝𝐚𝐭𝐚 - 𝐭𝐡𝐞𝐲'𝐫𝐞 𝐬𝐡𝐨𝐫𝐭 𝐨𝐧 𝐭𝐢𝐦𝐞.

Cogentiq Data Science serves as an agentic AI assistant for data scientists and analysts, streamlining the entire data science lifecycle. This includes exploring data, feature engineering, training, tuning, visualizing, and comparing models.

It provides instructions in plain language to receive reproducible code, interpretable charts, and comprehensive metrics, all in one platform.

Experience less switching, more clarity, and faster insights.

Learn how: https://t.co/qXvzJPKzs9

#Cogentiq #DataScience

[Rohit Joshi @rojoshi75]

It was a good to share our expereinces as well as learn on the challenges and opportunities that we can assist to deliver better citizen services. Thank you to Amazon Web Services (AWS) for co-hosting.

[Fractal @fractalai]

Exciting discussions at the recent 𝐀𝐈 𝐄𝐱𝐞𝐜𝐮𝐭𝐢𝐯𝐞 𝐑𝐨𝐮𝐧𝐝𝐭𝐚𝐛𝐥𝐞 hosted by @fractalai and @awscloud for the Victorian Public Services Department!

We explored the transformative power of AI in enhancing service delivery and productivity.

Key takeaways included:
🔸 The pressing need to boost AI literacy and awareness within the departments, to equip teams with a clear understanding of AI’s potential, limitations, and risks.
🔸 Small-scale pilots being crucial for demonstrating value and gaining practical insights to inform broader policies and adoption strategies.
🔸 Necessity of a robust data foundation and unified governance to deploy AI solutions responsibly and effectively to re-in force trust in AI-enabled public services.

Looking forward to applying these insights to our initiatives!

[Rohit Joshi @rojoshi75]

@fractalai @awscloud It was a good to share our expereinces as well as learn on the challenges and opportunities that we can assist to deliver better citizen services. Thank you to AWS for co-hosting.