Keshav Maheshwari

Something interesting is happening quietly across the legal industry. More lawyers are already using AI in their workflows than most people realize. But the biggest challenge isn’t capability anymore. It’s trust. Especially around: - confidential client data - internal documents - compliance expectations - where data is actually going I’ve been exploring privacy-first AI infrastructure and workflows for legal professionals, and I’d genuinely love to understand how lawyers are currently using AI in practice today. Not selling anything here; mostly listening and learning. If you’re experimenting with AI in legal work, feel free to DM me. Happy to also share some of the ideas and systems we’re building in beta.

The AI industry got it backward. We obsessed over model size while ignoring what makes AI useful: the orchestration layer. Utility = API (Reasoning) × Tool (Orchestration) A great tool makes a decent model feel magical. Stop chasing headlines; start engineering the system. That's where the real leverage lives.

@hiarun02 quality & reasoning

If DeepSeek V4 can do the same coding task for $5 why are people still paying $100 for Claude Code?

@tdinh_me Any plan to support Ollama Cloud models? @tdinh_me @TypingMindApp

TypingMind now supports 12 LLM providers natively 🥰 (you can chat with all of them in parallel at once!)

Interesting. dgld.ch is a solid institutional model. The Indian challenge is quite different though, we’re mostly dealing with 22k jewelry gold sitting in households with strong emotional attachment, not institutional allocated gold. When I worked on hGold, this cultural + redemption layer made things significantly more complex than pure custody models. Indian refineries stepping up would definitely help.

@nikhilkamathcio Had a brief opportunity to work on this with a Swiss client on a technical side, they already have this in another form as gold backed. dgld.ch. Its recently acquired by refinery MKS-PAMP. I believe the Indian Refinery have to step up on this and this can be done.

The world still runs on the dollar. But countries are quietly hedging out: buying gold, trading in non-USD pairs, building payment rails outside SWIFT. UPI has been incredible for India to say the least. To friends championing dollar-backed stablecoins, specifically dollar-backed, this seems like a bad idea long-term for India. Credit where due, to Modi govt and regulators, you got this one right in the face of a lot of pressure. On the other hand, if there were a gold-backed stablecoin and one could monetise the unutilised gold sitting in Indian households to return a yield… don’t know enough to talk about this, but thoughts?

Good point on XAUT and PAXG. Those work well for allocated/investment gold. The much harder (and bigger) opportunity in India is monetizing household jewelry gold (~95% of idle gold), where emotional attachment + physical redemption is a real issue. Custody is definitely the biggest blocker, we faced this heavily while working on hGold.

@nikhilkamathcio There already XAUT by @tether , and also PAXG, if you want gold backed stablecoin. And as of monetizing home gold, ig you would need to give up the custody which i don't think anyone specially indian household would be eilling to do, otherwise its a huge risk for issuer imo.

Good distinction. I worked on the second type you mentioned (monetizing household gold). The custody model is relatively straightforward, but getting Indian households to move 22k jewelry into a tokenized system is a completely different challenge; emotionally, logistically, and regulatorily. Curious which direction you think has a better shot in India.

@nikhilkamathcio On gold-backed stablecoins for India — the question splits into two designs. Pure custody (like Tether’s XAUT, ~$2.2B in allocated Swiss-vaulted gold): zero yield, 25 bps mint/burn fee, gold price exposure + liquidity. That’s it. The Indian opportunity is the second design.

Spot on about the 25,000 tonnes sitting idle. When I worked on hGold, this was exactly the opportunity we were chasing, but the hard part is getting households to trust a system with their emotional/jewelry gold and handling physical redemption at scale. Still believe it’s worth solving.

Dollar isn't dying. It's diluting. UPI proved India builds better when it builds sovereign. 25,000 tonnes of household gold. Zero yield. Zero monetisation. A gold-backed stablecoin doesn't export someone else's monetary policy. RBI got UPI right under pressure. Back them to get this one right too.

I worked on this. Co-led a project called hGold a couple of years back to design a physical gold-backed stable token for India. The vision matches what you described, but the real challenges are around jewelry gold, trust/custody, and physical redemption. Happy to share what we learned if useful.

Spot on. UPI + RBI’s firm stance on dollar stables was a smart sovereignty move. Gold-backed stablecoin though? That could be huge. Tokenize India’s massive idle household gold (~25k tonnes), give it yield via regulated vaults, and turn cultural savings into productive capital, without selling the heirlooms. Way better strategic fit than USD stables. Anyone working on this?

Agree. A few years ago I co-led hGold, trying to build exactly this. The biggest practical blockers we found were custody at scale, 22k jewelry redemption friction, and liquidity mismatch. Turning idle gold into productive collateral is powerful, but execution is complex. Would love to exchange notes.

@nikhilkamathcio Gold-backed stablecoins are interesting in theory, especially for India where households already hold massive amounts of idle gold. It could be really powerful if done right, almost like turning dead savings into productive collateral

Completely agree. A gold-backed stablecoin makes much more strategic sense for India than dollar-backed ones. A few years ago I co-led hGold, a project to design exactly this: a 1:1 physical gold-backed stable token built for Indian realities. We went deep into the practical problems most discussions skip: - 95%+ of household gold is 22k jewelry (emotional + refining issues at scale) - Custody, auditing & trust at national scale - Physical redemption friction - Liquidity mismatch The vision is powerful. Execution is where it gets hard. Huge respect for surfacing this @nikhilkamathcio. What do you see as the biggest solvable blocker right now?

AI’s 10x effect needs a base skill to multiply. Not a tool; an orchestrator. Your expertise is the input. Observed in my work, May 2026.

@CheekyCrypto The "Lender of Last Resort" play via Flowdesk is spot on. It’s not just treasury theater, it’s fixing real slippage on USDA/USDM and building the tight rails x402/AI agents actually need. Organic LPs alone move too slow for institutions. Clean and necessary move.

Analysis: The Institutional Liquidity Pivot On chain signals confirm the Cardano Foundation has begun an eight figure ADA liquidity deployment via Flowdesk. The target is a structural compression of slippage for USDA and USDM stablecoins. This isn't a simple treasury move; it's a targeted attempt to force institutional-grade depth. Per the Foundation's 2026 roadmap updates, this deployment utilizes Flowdesk’s market making infrastructure to stabilize the $219M+ TVL ecosystem. The second order effect is clear: By subsidizing liquidity for fiat backed assets like Mehen’s USDM, the Foundation is building the low-latency rails required for the recently integrated x402 protocol. s Cardano transitions into a verified x402 chain, the demand for deterministic, agent to agent payments is scaling. AI agents cannot tolerate the 2–3% slippage common in low cap DEX pools; they require the sub cent spreads typically reserved for CEX environments. This move signals a shift from passive governance to active market participation. By acting as a "Lender of Last Resort" for liquidity, the CF is effectively de risking the environment for the Draper Dragon Orion Fund and other institutional participants recently entering the ecosystem. Is this a necessary maturation of the ADA treasury or a sign that organic liquidity providers are failing to meet institutional demand? Option A: Strategic ecosystem hardening. Option B: Synthetic TVL preservation.

📊 USDT grabs bulk of $1.27B weekly growth, USDC down $362M. Tron liquidity dominant, ETH for regulated capital. Circle +13% on Clarity Act hopes. Aster fee-free USDC/USDT swaps (30 days). FDUSD +2.2% 24h. Regulatory clarity could spark USDC reversal? #Stablecoin #USDT #USDC

The future isn’t AI vs. crypto. It’s AI unlocking stranded energy for decentralized systems. Source: Live CBECI feeds, my hourly miner/grid correlation logs (May 1–10, 2026), POSOCO bulletins. #BitcoinMining #AIenergy #DePIN #StrandedEnergy #Web3Build #EnergyArbitrage

🛠️ What builders should actually do: - Use AI forecasting to tokenize renewable surplus timing (e.g., “1MWh wind power at 2 AM in Gujarat” as an NFT) - Build DePIN microgrids that auto-execute when AI-predicted surplus hits threshold (I’m testing this with a flared gas site, more soon) - Advocate for TOU tariffs rewards for flexible loads during verified renewable surplus windows → Stop fighting over energy. Start building the handoff layer.

Unpopular take: Stop defending Bitcoin’s energy use. Start noticing how AI’s grid pressure is solving its biggest criticism. Here’s my live field log (May 2026): No corporate links. No MSME research. Just what I’m tracking in CBECI + grid feeds.