[ NEURALARC_PRODUCT_—_OPENTHESIS ]
Valuation Terminal for Indian Equities
What it is
A stock terminal first and an AI product second — with one-click intrinsic valuation reports.
[ THE_PROBLEM ]
Why this existed
Retail investing tools show price and hide value. The intrinsic-valuation work that actually justifies a decision lives in analyst spreadsheets nobody else sees.
[ WHAT_WE_BUILT ]
What we built
A live market dashboard (indices with sparklines, gainers and losers, sector pulse, watchlist grid) and a per-company terminal with price history across 1M/6M/1Y/5Y, 52-week range, key ratios, and quarterly, P&L and balance-sheet tables. The centrepiece is a valuation report that encodes a classic intrinsic-valuation playbook as a schema: business story and life-cycle stage, explicit assumptions on growth, margins, reinvestment and cost of capital, a relative-multiples cross-check, bull/base/bear scenarios with probabilities rendered as a football field against the live price, and a margin-of-safety verdict. Made economical by a hardened cheap-model JSON pipeline — native structured output first, then balanced-block extraction, schema validation, one repair round-trip, and a provider fallback chain. All charts are hand-built SVG; no chart library.
- Live market dashboard with indices, sparklines, gainers, losers and sector pulse
- Per-company terminal with 1M / 6M / 1Y / 5Y price history and 52-week range
- Key ratios plus quarterly, P&L and balance-sheet tables
- Valuation report encoding story, life-cycle stage and explicit assumptions
- Growth, margin, reinvestment and cost-of-capital inputs made visible
- Relative-multiples cross-check against peers and history
- Bull / base / bear scenarios with probabilities rendered as a football field
- Margin-of-safety verdict against the live price
- Hardened JSON pipeline: structured output, balanced-block extraction, validation, repair, fallback chain
- Hand-built SVG charts with no charting library
[ HOW_IT_IS_USED ]
How a company uses it
Advisory firms and investing platforms give every client a documented valuation thesis instead of a price target with no working shown. The JSON-hardening pattern is reusable anywhere a cheap model has to produce reliable structured output at volume.
Built with
[ COMMON_QUESTIONS ]
Questions clients ask
Why does the JSON pipeline matter so much?
Because it is what makes the economics work. Cheap models are dramatically less expensive but less reliable at producing valid structured output. Rather than paying for a premium model on every report, the pipeline tries native structured output, falls back to extracting the first balanced JSON block, validates against a schema, attempts one repair round-trip, and finally fails over to another provider. That pattern is reusable anywhere you need reliable structured output at volume.
Why build charts by hand instead of using a library?
Control and weight. A football-field scenario chart plotted against a live price is not a standard chart type, and the common libraries add substantial bundle size for shapes we were going to customise anyway. For a handful of bespoke visualisations, hand-built SVG was the smaller and more precise option.
Is the valuation methodology sound?
It encodes a standard intrinsic-valuation framework — narrative to numbers to scenarios to price versus value — as an explicit schema. That is its strength and its limit: the framework is well-established, and the output is only as good as the assumptions, which is precisely why the assumptions are displayed rather than buried.
[ RELATED_WORK ]
Similar builds
Retail investing / fintech
AI Equity Research Platform
Generates full equity research reports, portfolio analysis and stock comparisons on demand.
Read case studyDebt capital markets
Bond Issuance & Investor Platform
Onboarding, deal workspace and investor portal for private debt issuance.
Read case studyCapital markets media & research
IPO & Market Data Platform
A public market-data destination fed by a fleet of automated ingestion jobs.
Read case studyIs this close to your problem?
Most engagements start with a version of something on this page. Tell us what is different about yours and we will tell you what it changes.
Start a conversation