Overall vision
Neolex is a project to bring regulations, policy and law into the modern age by transforming them from static, hard-to-navigate PDFs into living, structured, machine-readable infrastructure.
The ultimate goal is a world where regulations, circulars and policy documents exist as structured code with rich embedded context, on top of which an ecosystem of tools can be built.
Step 1 — SEBI PDF to HTML Converter
Foundation · In Progress
SEBI circulars are the first object of study because they are concrete, public, and painful. The converter takes a PDF circular and produces structured HTML carrying the circular number, date, subject, department, entity applicability, references, amendments and supersessions. This is not a cosmetic conversion. It is the first act of schema discovery.
Step 2 — Expand to Other Regulators
Foundation · Planned
Each regulator has its own document habits. Some publish HTML, some publish PDFs, some issue bilingual documents, and some publish material that looks closer to a court order than a circular. Neolex treats each source body as a separate input problem while forcing the output into a common structure.
Step 3 — Circular Database with Smart Filtering
Core Infrastructure · Planned
A database of circulars is not useful if it only searches titles. The text that matters often sits in the body of the document, and department labels are unreliable because regulated entities move across departments. The Neolex database treats body text as primary and metadata as the filter layer.
Step 3.5 — Git for Regulations
Core Infrastructure · Planned
Compliance questions are historical. What matters is not only what the rule says today, but what it said on a particular date. Git for Regulations maintains clean versions, dirty versions, amendment links, supersession history and point-in-time lookup.
Step 4 — RAG-Based Compliance Q&A Engine
Intelligence Layer · Planned
A serious compliance answer rarely lives in one circular. It depends on entity type, activity, size, date, master circulars, amendments and exceptions. The Q&A layer is not a generic chatbot; it is retrieval and reasoning over a structured regulatory corpus.
Step 5 — Compliance Tracking + Change Alerts
Entity-Facing Product · Planned
Compliance teams need to know what must be filed, when, by whom, and what changes when a circular is issued. The tracker converts regulatory obligations into operational workflows and updates them when the source law changes.
Step 6 — Bidirectional Compliance Checker
Regulator + Entity Product · Planned
A regulated entity preparing a submission and a regulator reviewing an application are doing the same underlying task from opposite directions: comparing a document against requirements. Step 6 is where the original Neolex vision becomes real.
Step 7 — Policy Impact Engine
Policy Intelligence · Planned
At this point Neolex moves upstream. Instead of only helping users comply with rules after they exist, it helps policymakers understand what a rule may do before it is finalised and what it did after implementation.
Step 8 — Live Feed + Academic Ecosystem
Platform Infrastructure · Planned
The corpus should not be a snapshot. It should watch regulatory websites, detect new material, route it through the right converter, update versions, and notify downstream documents when the rules they rely on change.
Step 9 — Peer Benchmarking and Best Practice Engine
Entity-Facing Product · Planned
Compliance is not exhausted by formal text. Public filings reveal emerging norms. The benchmarking layer compares a submission against both regulatory requirements and peer practice.
Step 10 — Enforcement Intelligence Engine
Intelligence Layer · Planned
Enforcement orders show law in action. Neolex can link orders, violations, penalties and tribunal reasoning back to the provisions they interpret, creating an enforcement history for each clause.
Step 11 — Cross-Jurisdictional Comparison Database
Research Infrastructure · Planned
This layer supports research, not compliance advice. It maps Indian rules to international counterparts while making structural differences explicit rather than pretending that every jurisdiction maps neatly onto another.
Step 12 — Regulatory Drafting Co-Pilot
Upstream Policy Tool · Planned
A drafting co-pilot checks a proposed circular against the existing corpus, flags ambiguous language, surfaces precedent, previews impact and warns where similar language has produced enforcement disputes.
Step 13 — Predictive Regulatory Intelligence Brief
Policy Intelligence · Planned
Consultation papers, speeches, enforcement spikes and international developments often point toward future regulatory movement. This is structured intelligence synthesis, not prophecy.
Step 14 — RegTech API Platform
Platform and Business Model · Planned
The capstone is infrastructure. RegTech startups, banks, law firms, researchers and regulators should not each rebuild the regulatory corpus from PDFs. They should build on a structured layer.
Footnotes and peripheral ideas
Some ideas are deliberately not central to the roadmap: a flashy visual dependency graph, a consultation-paper response generator, a rupee-level compliance cost quantifier and a separate plain-language retail layer. The core build remains the structured regulatory infrastructure layer.
The full idea is intentionally public. The value is in execution, not secrecy.