By Robert Fon / August 21, 2024 (Updated October 21, 2026)
If your account research still means opening ten company websites and checking the page source for a script tag, you are only seeing the technologies a browser can load, which misses most backend tools, data warehouses, and anything mentioned only in a job posting. A technology detection API fixes that by returning structured, source-linked records of which technologies a company shows evidence of using, built from the same five public signals PredictLeads tracks across 99M+ companies. This guide covers what a technology detection API is, the two ways you can query it, the fields it returns, who uses it, and how to turn “find companies using a specific technology” into a clean, filterable list instead of a pile of browser tabs.
TLDR:
- A technology detection API returns structured, timestamped records of which technologies a company shows evidence of using, built from public sources instead of a single page scan.
- You can query it two ways: forward, from a known company to its technologies, or reverse, from a technology name to every company showing evidence of using it.
- A detection is evidence that a technology appears in a public signal, such as a script tag, a DNS record, or a job description. It is not confirmation that the tool is actively running today.
- Most single-source scanners only read what loads in a browser, so backend systems, data warehouses, and many CRM deployments surface mainly through job descriptions and DNS records instead.
- PredictLeads Technology Detections cover 2B+ detections since 2018 across 99M+ companies and 240,000+ technologies, each record carrying
first_seen_at,last_seen_at, and abehind_firewallflag.
What Is a Technology Detection API?
A technology detection API is a programmatic interface that returns structured records showing which technologies a company’s public footprint provides evidence of using, instead of requiring someone to check a website by hand. Each record ties a company to a technology with a timestamp and a source, so a sales team, a data pipeline, or an AI agent can ask either side of the question: what does this company use, or who uses this technology.
The PredictLeads Technology Detection API answers both, drawing on website script tags, DNS records, IP ranges, cookies, and job descriptions, five public evidence types that each catch a different part of a company’s stack. Technology detection is the mechanism behind what PredictLeads calls technographic data, and PredictLeads has logged 2B+ of these detections since 2018, covering 240,000+ tracked technologies across 99M+ companies.
What a Technology Detection API Is Not
A technology detection API is not a website scanner that only reads what loads in a browser. Script tags and page code catch frontend tools such as analytics and chat widgets, but they miss a data warehouse, an ERP system, or a CRM a company’s sales team adopted last quarter.
A detection also is not confirmation that a tool is installed and running today. It is evidence that the technology appeared in a public signal as of its last_seen_at date. If a detection stops appearing, the fair framing is that the technology is “no longer detected,” since a script change, a recrawl gap, or a signature update can all cause the same gap without the company actually removing anything.
Finally, a technology detection API is not a contact or people database. PredictLeads records what a company shows evidence of using at the company level, not personal emails or phone numbers.
Two Ways to Query a Technology Detection API
A technology detection API answers two distinct questions, and knowing which one you are asking determines the endpoint you call.
Forward Lookup: Company to Technologies
Start with a company domain when you already have an account in mind, for example during enrichment or ahead of a renewal conversation. The PredictLeads endpoint GET /companies/{company_id_or_domain}/technology_detections returns every technology that company shows evidence of using, each with its own first_seen_at and last_seen_at. For the full walkthrough of this pattern, see how to detect a company’s technology stack.
Reverse Lookup: Technology to Companies
Start with a technology name when you want a list of accounts instead of one profile, for example to find every company that shows evidence of using a specific tool. The endpoint GET /discover/technologies/{technology_id_or_fuzzy_name}/technology_detections, ordered by first_seen_at descending, returns companies tied to that technology. For the workflow on filtering that list by recency and expanding it with lookalikes, see how to find every company using a technology by name.
| Query Direction | You Start With | Endpoint | Best For |
|---|---|---|---|
| Forward | A company domain | GET /companies/{domain}/technology_detections |
Account research, renewal timing, enrichment |
| Reverse | A technology name or ID | GET /discover/technologies/{technology_id_or_fuzzy_name}/technology_detections |
Competitor displacement, integration targeting, market sizing |
A third endpoint, GET /extended_technology_detections/{id}, pulls the full detail behind a single detection, including which subpage, job opening, or DNS record it came from.
How the PredictLeads Technology Detection API Works
Authentication and Requests
The API sits at https://predictleads.com/api/v3 and supports three authentication methods: HTTP headers (X-Api-Key and X-Api-Token, the recommended method), HTTP Basic Auth, or query parameters. PredictLeads also supports OAuth2, including an OAuth 2.0 Client Credentials flow for server-to-server calls. Credentials are issued on the subscription page after you sign up and verify your email. Responses follow the JSON API specification, and the rate limit is 60 requests per second, with a 429 response and a Retry-After header if you exceed it.
Key Response Fields
Every detection carries first_seen_at and last_seen_at timestamps, a score, and a source_count showing how many independent signals back it. The behind_firewall field is a boolean: false means the detection came from something visible on the company’s website, such as a script tag or HTML. Extended Technology Detections add detection_source and detection_source_type, for example script_src, cookies, headers, or dom, plus department_onet_codes when the source is a job description.
If a technology stops showing up in new detections, treat it as no longer detected rather than removed. A missing signal can mean a script changed, a recrawl has not run yet, or a fingerprint updated, not that the company dropped the tool.
Example Technology Detection API Request and Response
A reverse lookup for a specific technology looks like this:
curl --request GET
--url https://predictleads.com/api/v3/discover/technologies/salesforce/technology_detections
--header 'X-Api-Key: {your_api_key}'
--header 'X-Api-Token: {your_api_token}'
Each result in the response ties a technology record to a company record, with attributes shaped roughly like this (simplified for readability):
{
"technology": { "name": "Salesforce", "category": "sales" },
"company": { "domain": "example.com", "name": "Example Inc." },
"first_seen_at": "2026-02-11",
"last_seen_at": "2026-10-03",
"behind_firewall": true,
"source_count": 2,
"score": 0.91
}
A behind_firewall value of true here means the detection came from a source other than the public website, most often a job description naming Salesforce as a required skill, which is why enterprise CRM, data warehouse, and ERP tools show up more often behind the firewall than in page code alone. Paginate with the page parameter for lists beyond the first page, and the discovery endpoint typically uses about 1 credit per record returned. You can also test requests directly in the SwaggerUI playground before writing any code.
Who Uses a Technology Detection API
Four groups query technology detection data for different reasons:
- GTM and sales teams: build prioritized lists for competitor displacement, integration-fit targeting, and personalized outreach.
- Developers and data teams: pipe detections into a lead-scoring model, a CRM enrichment job, or a data warehouse table.
- Market intelligence and competitive intelligence analysts: track adoption trends across a category or a competitor’s customer base over time.
- AI agents: query the same data conversationally through MCP, the pattern covered in our comparison of company data APIs for AI agents, for example asking an agent to “find 10 companies using HubSpot” instead of writing a custom integration.
Technology Detection API Use Cases
A reverse lookup by technology name supports several concrete plays:
- Competitor displacement: find companies showing evidence of using a competitor’s product, then prioritize by how recently the detection first appeared.
- Integration-fit targeting: find companies using a tool your product integrates with, so the pitch starts from a real technical fit.
- Account scoring: feed technology detections, alongside Job Openings and News Events, into a scoring model instead of relying on a single signal.
- CRM and warehouse enrichment: attach technology fields to existing account records on a schedule instead of researching one company at a time.
- Market adoption tracking: watch how many companies in a category show evidence of adopting a technology over a given period.
Why Multi-Source Detection Matters
Script tags alone miss most of what matters to a sales team. A content management system or a chat widget loads in the browser, but a data warehouse, an ERP system, and plenty of CRM deployments run entirely behind the login screen, where a scanner never reaches. PredictLeads closes that gap with five independent sources: website script tags, DNS records, IP ranges, cookies, and job descriptions. For the full breakdown of each method, see how technographic data is collected.
Job descriptions carry the most weight for enterprise tools specifically. A company hiring a “Snowflake-certified data engineer” is stronger supporting evidence of a Snowflake deployment than a public script could ever be, since the tool itself produces no public page code. Combining sources also raises confidence: a detection backed by both a DNS record and a job description, reflected in source_count, is harder to explain away than either signal alone.
Technology Detection API Delivery Methods
The data behind a technology detection API does not have to arrive one request at a time.
| Delivery Method | How It Works | Best For |
|---|---|---|
| API | On-demand requests against company and discover endpoints | Real-time lookups, enrichment at request time |
| Flat files | Bulk JSONL delivery to S3, Google Cloud Storage, or SFTP | Warehouse loads, backfills, large-scale analysis |
| Webhooks | A push notification when a new detection appears for a followed company | Trigger-based alerting workflows |
| MCP | Natural-language queries through mcp.predictleads.com | AI agent workflows, conversational data access |
Flat files and webhooks are arranged through the sales team, while the API and MCP are self-serve. For a full comparison of when each method fits a given workflow, see company data API vs. flat files vs. webhooks vs. MCP.
How PredictLeads Delivers Technology Detection Data
PredictLeads has logged 2B+ Technology Detections since 2018, covering 240,000+ tracked technologies across 99M+ companies in 195 countries. Every record carries first_seen_at, last_seen_at, a score, and a source_count, and Extended Technology Detections add detection_source, detection_source_type, and the subpage, job opening, or DNS record each detection came from, so nothing arrives as an unexplained flag.
The same API that returns technology detections also returns Job Openings (279.2M+ records since 2018), News Events (10M+ since 2016), Financing Events (210,800+ since 2016), and Connections (371.8M+ since 2019), so a technology signal does not have to stand alone. A company that shows evidence of adopting a new data platform while hiring for it in the same quarter is stronger supporting evidence than either dataset checked alone.
Delivery runs through API, flat files, webhooks, and MCP, and every dataset is collected from public sources only, under SOC 2 Type II certification and GDPR and CCPA compliance.
Final Thoughts on the Technology Detection API
A technology detection API only pays off when you treat a detection as evidence, not a certainty, and when you query it from whichever direction the question actually starts: company first, or technology first. Pair it with the hiring, funding, and news signals already in the same API, and a single detection becomes part of a timed account list instead of an isolated data point. If you are comparing this approach against a single-source scanner, see how PredictLeads compares to Wappalyzer, BuiltWith, SimilarTech, and TheirStack, or see the full field-by-field case for the category in our guide to the best technographic data providers.
Ready to see this in your own data?
Get 100 free API requests per month – no credit card, no sales call. Start free →
Frequently Asked Questions
What is a technology detection API?
A technology detection API is a programmatic interface that returns structured, timestamped records of which technologies a company’s public footprint provides evidence of using. Instead of opening a website and checking the page source, you call an endpoint and get a company-to-technology or technology-to-company list back as data. PredictLeads builds its Technology Detection API from five sources, website script tags, DNS records, IP ranges, cookies, and job descriptions, across 2B+ detections logged since 2018.
How do I find companies using a specific technology with an API?
Call the reverse-lookup endpoint, GET /discover/technologies/{technology_id_or_fuzzy_name}/technology_detections, with the technology’s name or ID, and the response returns companies tied to that technology, ordered by first_seen_at descending. This is the pattern for competitor displacement and integration-fit lists, since you start from the technology rather than a single account. PredictLeads covers this across 240,000+ technologies and 99M+ companies.
Does a technology detection mean a company is actively using that tool right now?
Not by itself. A detection means the technology appeared in a public signal, such as a script tag, a DNS record, or a job description, as of its last_seen_at date, which is evidence of use rather than confirmation of an active, running install. If the detection stops appearing, the accurate description is “no longer detected,” since a script change, a recrawl gap, or a signature update can produce the same gap without the company removing the tool.
What is the difference between a technology detection API and a website technology scanner?
A website scanner, including most browser extensions, reads only what a page exposes in its code, headers, and cookies, which covers frontend tools well but misses backend systems with no public footprint. A technology detection API that also reads DNS records, IP ranges, and job descriptions catches enterprise tools, such as a CRM or a data warehouse, that a scanner alone would miss. PredictLeads flags this directly with a behind_firewall field on every detection.
How much does the PredictLeads Technology Detection API cost in 2026?
PredictLeads offers 100 free API requests per month with no credit card required, and paid usage runs on a pay-as-you-go, volume-based model with a $40 minimum monthly charge once you exceed the free tier. The discovery endpoint used for reverse lookups typically costs about 1 credit per record returned. Flat files, webhooks, and enterprise volume pricing are arranged through the sales team.
Pingback: Lead Scoring Models Oct 2026
Pingback: Technographics Dataset: Tech Stack Leaders | October 2026