Verifies the mandate
Every asset receives a permanent AIS Asset Key. Principal authority is documented before anything is shared, and duplicate or competing mandates are detected automatically.
One governed rail from mandate to close: verify who holds authority, diligence what matters for the asset class, score it, place it with the right capital under NCNDA, protect the fee, and learn from every outcome. Run by one accountable human and a governed team of specialised AI agents, openly disclosed.
The AIS Platform is a governed deal rail for off-market real assets. It verifies principal authority, diligences title and entitlements, scores each deal, matches it to capital on stated and observed behaviour, protects the fee with NCNDA-gated rooms, and learns from outcomes.
AIS does not add AI to real estate advisory. It rebuilds the advisory process around verification, evidence and governance, and uses AI to run it at institutional standard.
Every asset receives a permanent AIS Asset Key. Principal authority is documented before anything is shared, and duplicate or competing mandates are detected automatically.
AI analysts extract and tag every data point by provenance, investigate title, entitlements and counterparties, and score the deal with the AIS Deal Score™.
An independent risk function stress-tests every locked score and holds veto power.
Matching weighs what capital partners actually do, captured from every teaser view, data-room session, question and pass, alongside what they say they want.
NCNDA-gated data rooms, per-recipient watermarking, logged access, pre-agreed fee splits and circumvention monitoring are built in.
Post-close performance feeds back into scoring calibration, so the framework is measured against realised results, not opinion.
Agents act on events (a deal submitted, a score locked, a risk review cleared), not on a rigid chain.
Deal state transitions, score immutability, document access tiers and data visibility are enforced by the database itself, not by agent discretion.
Every agent action is logged with its reasoning, data used, confidence and escalation status. The log is append-only.
Each action is scored for external exposure, capital impact, reputational sensitivity and doctrine alignment, and routed accordingly.
A system-wide freeze halts all external-facing activity while governance and audit continue.
Unverified mandates can move through intake and scoring for internal learning, but cannot enter matching. The rule is enforced at the database level.
The asset is real and uniquely identified. The party presenting it has verifiable authority. AIS is not one of several intermediaries circulating it.
Every asset receives an AIS Asset Key. Identifiers are recorded against it, so every future deal on the same asset attaches to the same key and transaction history accumulates.
Each submission is matched against the registry on hard identifiers (licence number, khasra, RERA ID, CIN) and on location plus asset class. A match opens a conflict record showing both chains of custody and the price difference.
Conflicts with a price gap above 15% are treated as reputation-sensitive and are not distributed until resolved.
Deals arrive as teasers, information memoranda, scanned documents, spreadsheets, emails and forwarded messages. AIS turns them into a structured Deal Brief.
From the Partner Portal, a dedicated deals inbox and forwarded messages. Asset-class detection selects the right field schema: land, hospital, hotel, industrial, lease or commercial.
Every extracted field carries its value, a confidence level and a page or cell reference to its source.
Corrections are logged and used to improve extraction accuracy.
Every document processed, including deals AIS passes on, enriches the AIS comparables database.
Every data point is classified at entry. Extraction never upgrades provenance; only verification does.
If more than 40% of a deal’s parameters rest on claimed or unknown data, Risk is automatically notified. Above 60%, the deal returns to Research with specific verification requests before it can be scored.
For Indian land and institutional assets, entitlement status is often the largest value driver and the largest risk. Each asset carries a register with authority, reference, status, validity, conditions and outstanding dues. For pending entitlements, the engine estimates probability and timeline, clearly labelled as estimates.
Principals, brokers and capital partners are checked for identity, beneficial ownership via company filings, sanctions and PEP status, adverse media and litigation. High-risk parties cannot receive introductions without governance sign-off. Checks refresh annually or on change.
For value-add and repositioning deals, renovation or completion capex is estimated from drawings, site imagery and benchmark rates, as a low / base / high range with stated assumptions. Indicative, and never a substitute for a contractor BOQ.
Integrates 360° capture into the data room, with a structured condition report, so overseas capital partners can assess an asset without travelling.
A land parcel is not a hotel. Every score combines a Common Core that applies to all deals (50%) with an asset-class module (50%).
Covers offices, schools, PGs and operating senior living assets.
Covers warehouses, logistics parks and industrial buildings. Vacant industrial land is scored under the Land module.
The scoring analyst does not see who the potential buyers are. The risk analyst reviews the locked score independently and asks one question: what could go wrong?
Score immutability is enforced by the database. No role can alter a locked score.
The revenue function cannot influence scores and has no authority over scoring or risk staff.
Disagreements between revenue and investment functions are arbitrated by the Chief Governance Officer, who reports directly to the CEO.
Predicted scores against realised outcomes report to Governance, not to Scoring, so the framework never validates itself.
Vetoed deals that later prove sound count against the risk function’s record.
AIS Connect replaces email and message-based teaser distribution with a controlled feed inside the Partner Portal. Verified, scored and risk-cleared mandates are pushed to matched capital partners, and every interaction becomes data.
Each match combines stated preferences (geography, asset class, ticket size, structure, risk appetite) with behavioural fit (response speed, deals progressed and declined, actual risk tolerance, drop-off points, the stated-versus-actual gap). Behavioural weight rises as evidence accumulates.
New capital partners start on stated preferences only. Every match is proposed with written reasoning. Capital partners see a score band, not the exact score, and never see other investors’ information.
Parties transacting directly after an intermediary’s introduction is the most common way intermediaries lose their fee. AIS builds the protection into the platform.
Thesis, asset, market, entitlements, financials, Deal Score breakdown, risk review and open diligence items. Every factual claim carries a source reference and provenance tag; unsourced claims are not permitted. Memos are version-controlled and locked with the score.
After introduction, milestones, documentation (NDAs, term sheets, LOIs, closing documents), counterparty communication and stalls are tracked. Answers to investor questions are drafted from the data room with source references and approved before release.
For JV, under-construction and development-finance placements, periodic site capture and contractor reports are compared against schedule to verify claimed progress before capital tranches are released.
Capital partners can opt into quarterly reporting on closed deals: occupancy, revenue, NOI or EBITDA, capex against plan and entitlement milestones. The data closes the calibration loop.
Leading indicators. AIS does not wait twelve months to learn. Investor engagement rate, diligence progression, pricing renegotiation, time to close and pass-reason clusters are tracked from the first week and reviewed against scoring.
AIS is structured like an institutional firm. Each seat has a defined mandate, reporting line, data permissions, performance metrics and escalation rules. The model is openly disclosed: AI team members do not claim to be human, and every material capital allocation decision is governed by human oversight and an independent governance framework.
The Governance function is independent of every operating department, reports directly to the CEO and cannot be overridden by any other executive.
Agents earn autonomy through demonstrated judgement, not tenure. Advancement is performance-gated, per agent and per capability.
Never delegated at any phase: amending the Doctrine, disabling governance or the kill switch, changing agent identity, an agent changing its own autonomy level, or acting in the CEO’s name without approval.
Pipeline status for your own submissions, information requests and messages. Never scores, risk reviews or match reasoning.
Curated teasers (asset class, city, ticket range, structure, score band), then the data room after NCNDA, then the full package after an approved introduction. Never exact addresses before NCNDA, and never other investors’ information.
AIS provides research, information and introduction services. It does not provide investment, legal or tax advice. Every party should conduct its own diligence and take independent advice.
Permanent IDs and chains of custody. Every repeat deal adds history.
Entitlements, title, filings and litigation joined on one asset key.
What each investor views, asks, signs and passes on. No listing site sees this.
Predicted scores measured against realised outcomes, under independent governance.
The AIS Platform is a governed deal rail for off-market real assets. It covers mandate verification, document intelligence, diligence, the AIS Deal Score, independent risk review, capital matching, fee protection, transaction management and post-close reporting. Access is by invitation.
The AIS Deal Score is a modular, confidence-weighted score. A Common Core that applies to every deal makes up 50%, and an asset-class module for land, hospital, hotel, industrial, lease or commercial assets makes up the other 50%. Each parameter is scored 1 to 10, adjusted for data confidence, and the score is locked before matching.
Every asset receives a permanent AIS Asset Key with its identifiers recorded. Principal authority is classified as a direct principal letter, board resolution, power of attorney, attested broker chain or unverified. Duplicate and competing mandates are detected automatically, and unverified mandates cannot enter matching.
It classifies every data point as verified, inferred, claimed or unknown, with scoring weights of 1.0, 0.8, 0.5 and 0.3. Extraction never upgrades provenance; only verification does. If more than 40% of a deal’s parameters rest on claimed or unknown data, Risk is notified.
Through an NCNDA gate, tiered documents, per-recipient watermarking, an append-only access log, fee splits agreed in advance, and monitoring of company filings, sale-deed records and data-room activity during the tail period.
No. AI agents prepare the work, but every introduction is approved by a human, and an independent Governance function can override any other department. AI team members do not claim to be human, and the model is openly disclosed.
Mandate holders see pipeline status for their own submissions, never scores, risk reviews or match reasoning. Capital partners see curated teasers with a score band, then the data room after NCNDA, then the full package after an approved introduction. They never see other investors’ information.
Access is granted by invitation, in cohorts. Request access and tell us which seat you would take; we grant access when the platform is ready for your profile.
Plain-language definitions of the Indian real estate, healthcare and hospitality terms used across the AIS Platform. See the full glossary.
Tell us which seat you would take. We review every request and grant access when the platform is ready for your profile.
Request access