DINQ: An AI Agent Engine for Finding and Analysing AI Talent

Hiring in AI is hard because the talent is scattered. The engineers and researchers who have actually trained large models, built multimodal pipelines or worked on inference optimisation leave their traces in preprints, repositories and conference talks, not in job-board profiles. DINQ approaches the problem with AI agents: locate candidates across public sources first, then explain in structured form what makes each one strong.

DINQ is a talent engine focused on the AI field. Its agents identify researchers and developers from public sources and help companies and research institutions shortlist people who fit a specific technical need. It is not another CV database; it converts public technical traces into signals a hiring team can actually judge.

The data layer consolidates several authoritative public sources, Google Scholar and GitHub among them. Academic output, code activity and stack preferences are merged into one candidate record used to assess research record, engineering ability and hands-on experience together.

Beyond the web interface, DINQ offers DINQ-as-a-Service (DaaS), a model that can be embedded into an existing recruiting stack so search and profiling run inside the workflow a team already uses.

What DINQ does

Global talent search

Agents locate AI specialists across public sources. The question it answers is not who applied, but who is genuinely working on the problem, which matters when a team has to build a candidate pool from scratch.

Multi-source consolidation

Records combine scholarly and code signals so a strong publication history is not mistaken for production experience, and an active repository is not mistaken for systematic research.

Semantic search

Besides a name or a scholar profile URL, you can describe the profile you need in plain language, for example a researcher focused on multimodal models with product experience. Ranking considers technical background, not just keyword overlap.

Fine-grained filters

Stack, years of experience, location and salary expectation can be combined to shrink a broad pool to a list that can be reviewed person by person.

Deep candidate insight

Each profile can be expanded into a multi-dimensional analysis with a chosen depth, surfacing key attributes and potential rather than a chronological CV.

Visual talent profiles

A skill radar, a growth timeline and an influence score show the shape of someone's ability and the path of their career at a glance.

Collaboration graphs

The platform maps academic co-authorship, industry connections and how influence propagates, which helps judge how wide a candidate's working network really is.

Live trend insight

Aggregated analysis covers hot technical directions, talent-flow heat maps and salary references by subfield.

Event-driven updates

Profiles are tracked continuously, so a new paper or repository is reflected instead of the record freezing on the day it was created.

DaaS integration and demos

Enterprises can connect the service to their own recruiting system; a demo request lets a team validate fit before committing.

How to use it

1. Create an account and confirm how seats and permissions are split across the team.
2. Start from one concrete need: paste a scholar profile URL, type a name, or write a natural-language description of the profile you want.
3. Narrow the result set with filters on stack, experience, location and salary expectation.
4. Open the deep insight view for each shortlisted person and compare the skill radar with the growth timeline.
5. Check the collaboration graph when team fit and network reach matter.
6. Calibrate compensation using the salary references before the first conversation.
7. For enterprise use, request a demo and confirm integration scope and data boundaries for DaaS.

Who it fits

• Recruiters building an AI team who need candidates beyond the inbound CV pool.
• Research labs looking for collaborators in a narrow technical direction.
• Hiring managers who want a technical read on a candidate before the first call.
• Investors and analysts tracking where AI talent moves and what subfields pay.
• Researchers and engineers who want an outside view of their own position.

Limits to know first

• No public pricing, plan structure or free tier is disclosed, so budget has to be confirmed with the vendor.
• Coverage depends on public data; internal projects, proprietary systems and non-English communities are represented unevenly, so scores are signals rather than verdicts.
• Only a few sources are named publicly, and the full list, refresh cadence and historical depth are not documented.
• The formulas behind the influence score and skill radar are not published, which makes cross-domain score comparison unreliable.
• Because personal data is processed, source legality, retention and deletion terms need checking against local privacy rules.
• Semantic search quality tracks the quality of the query; vague descriptions return vague pools.

FAQ

How is this different from a job board?

A job board shows who is looking for work. DINQ shows who is doing the work, drawing on public research and code, so passive candidates are reachable.

Do I need a candidate name to start?

No. You can begin from a scholar profile URL or a natural-language description of the profile you need.

Can the profile decide a hire?

It should support screening, not replace it. Interviews, collaboration history and team fit are outside what public signals can show.

Do we have to replace our ATS?

No. DaaS mode connects search and profiling into the recruiting system you already run.