Browse documentation
PMS & AIF Desk
Portfolio Management Services (PMS) and Alternative Investment Funds (AIFs) rely on multiple sources for research: news feeds, corporate filings, earnings calls, investor presentations, social media, and internal analyst notes.
Research teams often spend hours every day reviewing filings across 200-300 companies, trying to determine which filings are actually important, what changed, and whether the information affects their investments.
The challenge is not access to data. The challenge is identifying meaningful signals quickly.
Many listed companies release raw audio recordings without transcripts or structured summaries. Analysts then have to listen through long calls to extract management commentary, guidance changes, operational risks, or early indicators that may affect market sentiment.
Drishti is built as a toolbox for that research layer. Each tool solves a specific research problem. Together, the tools help teams see how news, filings, earnings, concalls, and market activity interact with price movement and portfolio exposure.
When a company releases an important filing or management gives new guidance during a concall, Drishti can help highlight the change immediately. When combined with market price data inside a research stack, teams can also track how the stock reacts after the event.
Rather than spending months building custom pipelines to collect, parse, normalize, and interpret market events, research teams can plug into Drishti's API layer and receive structured data ready for analysis, alerts, and agent workflows.
Where Drishti fits
| Research need | Drishti surface | What the team gets |
|---|---|---|
| Corporate filings and announcements | `/v1/announcements` | Filtered filing records by symbol, category, date, and detail level. |
| Earnings updates | `/v1/earnings` | Structured earnings filing context for quarterly result monitoring. |
| Concall analysis | `/v1/concalls` | Conference-call records, transcript access, and management-commentary workflows. |
| News and sentiment signals | `/v1/news` | Symbol-scoped news that can be combined with price data and internal portfolio context. |
| Real-time monitoring | WebSocket streams | Live feeds for news, block deals, announcements, earnings, concalls, and alerts. |
| Agent research workflows | Drishti MCP | Tool-call access so AI agents can retrieve the right market context directly, including prices, sectors, movers, and 52-week discovery. |
Announcements and corporate filings
/v1/announcements returns corporate filings filtered by symbol, category, or date. Instead of manually checking portals, sorting filings, prioritizing items, and assigning work across hundreds of companies, teams can query only what is relevant to their portfolio.
The response is structured for analysis or alerting, so a system can route filings by category, issuer, source timestamp, portfolio match, and materiality.
Earnings and concalls
/v1/earnings helps teams monitor structured earnings updates. /v1/concalls supports conference-call workflows, including transcript and recording access when available.
Together, these endpoints reduce the manual work of tracking result cycles, finding management commentary, and spotting guidance changes or operational flags. Analysts can move from listening and copying to reviewing structured evidence.
News and sentiment signals
/v1/news returns news by symbol. When combined with internal price data, teams can correlate news events with market reactions after an event has occurred, or power near-live monitoring through WebSocket streams.
This avoids the need to build a custom news ingestion pipeline just to create a usable research feed.
MCP for research agents
Drishti also exposes market data tools through MCP, which lets AI agents call Drishti directly instead of waiting for developers to build every retrieval path by hand.
In practice, an analyst or an internal research agent can ask questions such as:
Check recent filings, earnings updates, concall commentary, and news for the companies in this portfolio. Highlight material changes and cite the data source for each point.With MCP, the agent can choose the right Drishti tools for the task: fetch announcements, search concalls, inspect earnings, pull news, check prices, compare sector context, inspect top movers, or resolve symbol metadata. The agent does not need raw database access, and the team does not need to keep uploading documents into each chat.
Use MCP when the workflow is conversational, exploratory, or agent-driven. Use REST APIs and WebSockets when the workflow needs deterministic backend control, scheduled jobs, dashboards, alerts, or persistent product features.
This gives research teams two paths from the same data layer: engineers can build production workflows with APIs and streams, while analysts and agents can investigate market questions through MCP tool calls.
Real-time research systems
News, announcements, earnings, and concalls are also available through WebSocket streams. This makes it possible to build real-time analytic systems without maintaining custom ingestion pipelines for every source.
A PMS or AIF team can stream events into watchlist monitors, desk alerts, internal research queues, or agentic systems that prepare first-pass summaries for human review.
Why use Drishti instead of building custom
Building these workflows in-house means solving data sourcing, parsing, normalization, and delivery for every filing type, audio format, and broken PDF. That is months of engineering work, followed by ongoing maintenance.
Drishti handles the hard parts so teams can ship the research workflow in days. It is built with agentic workflows in mind, so engineers do not have to spend hours turning documentation into retrieval code before a system can become useful.
Our open-source stack provides a practical playground for trying the APIs and understanding the kinds of workflows that can be built on top of Drishti.