> ## Documentation Index
> Fetch the complete documentation index at: https://guide.kello.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Write a great brief for Kello

> A great Kello brief is your JD plus location, work setup, budget, must-haves vs nice-to-haves, companies to target or avoid, and people you'd hire.

A great Kello brief is your job description plus the few things it doesn't say: location, work setup, budget, which requirements are must-haves and which are nice-to-haves, companies to hire from or avoid, and a few LinkedIn profiles of people you'd hire in a heartbeat. Those few lines decide most of what lands in your first list.

You don't need a polished JD (Job description: the document that describes the role.). Pasted text, a PDF, a Word doc or a link to the job post all work. No JD? Send the role name and a few lines on who you need.

## What should a brief include?

* [ ] **Role and company in the subject line.** See [the subject line](#what-should-the-subject-line-say).
* [ ] **The JD**, with responsibilities, not only requirements. Required
* [ ] **Location and work setup.** City, office days, or remote with a time zone. Kello asks if missing
* [ ] **Budget.** A range, or `no fixed budget, use market rates`. Kello asks if missing
* [ ] **Experience band.** For example, 8–10 years.
* [ ] **Must-haves vs nice-to-haves.** What truly matters, and what can bend.
* [ ] **Companies to hire from or avoid.**
* [ ] **People you'd hire in a heartbeat.** A few LinkedIn profiles, with a line on why. See [how to share example profiles](#how-do-i-share-example-profiles). Optional
* [ ] **Your website**, if your company name is shared with others.

For a fill-in-the-blanks version, use the [general template](https://guide.kello.ai/kello/brief-templates.md#general-template-for-any-role).

## What holds up the first list?

Most delays happen before the search starts. The three things Kello asks about most are budget, location and work setup. Briefs that include all three usually start within minutes, with no questions. Briefs without them lose at least one round trip.

**Real example: same person, two roles.** A content creator sent two roles. The first was a PDF and nothing else. Kello replied: *"Before I can start sourcing, could you let me know the budget or salary range…"* The second email led with the details: budget up to ₹50,000 a month, full-time, on-site in a small city in North India, 2–5 years of experience. Kello started on it within 3 minutes.

## Do I need to give a location and budget?

Yes. If either is missing, Kello asks one short question before searching. Putting both in the first email saves the wait.

```text Reply to Kello wrap theme={null}
Budget: No fixed budget, use market rates
Location: Bangalore, 3 days a week in office
```

**Budget.** You don't need a number. A range is a complete answer, and so is `no fixed budget, use market rates`.

> [!WARNING]
>
> "Flexible" isn't a budget. Kello treats it as unanswered and asks again. Give a range, or say "no fixed budget, use market rates".

Check the budget with the hiring manager before you send it. Kello uses it to calibrate seniority, so a guess that turns out wrong costs a full round (Each list is a round: one list, plus your marks and feedback on it.).

**Location and work setup.** Give the city and whether the role is on-site, hybrid (how many days) or remote. For remote roles, add the time zone. Say whether people willing to relocate are welcome. Without it, people can drop out late over relocation.

A role open in two cities is one role: list both locations and the list stays up to 20 people. Want a full list for each city? Send each city as a separate role.

## What should the subject line say?

The role title and your company, for example `Backend Architect – [Your company]`. Kello knows which role and which company the thread is about from the first line, and your team can find the thread when they join it.

Keep one role per thread so the subject stays true.

## Which JD formats work best?

| Format | How well it works | Tip |
| - | - | - |
| JD text pasted into the email body | Works best | The most reliable way to start fast. |
| PDF or Word attachment | Usually works | Attach it as a backup even when you paste the text. |
| LinkedIn, ATS (Applicant tracking system: the software your team uses to track candidates.) or careers-page links | Usually works | Paste the key details too, in case the link opens a general careers page. |
| Notion links and some job-board links | Often fails | Paste the text, or attach a PDF or Word file. |

**Forwarding a hiring post from social media?** Kello can start from it if the post names the role, company, location and pay. Most don't, so add a line: `Budget: market rates. Location: remote, US time zones.`

Before you send, check you attached the right JD. A JD for a different role gets sourced as that role.

## What must the JD include?

What the person will do, not only what they need to have. A brief that lists only requirements and target companies can't start, because Kello has no role description to search for. A few lines of responsibilities fix it.

## Why split must-haves from nice-to-haves?

Most JDs list ten requirements with equal weight. Kello needs to know which two or three are firm.

Firm requirements stack, and each one narrows the pool. When not enough people meet the bar, Kello says how many cleared out of how many it evaluated, and which requirement is narrowing the search. If you've already said what can bend, Kello can make a better call.

If a list comes back short, ask what constrained it, then reply with the one constraint that can bend:

```text Example reply: what can bend wrap theme={null}
Open to people willing to relocate, and up to 35 LPA.
```

LPA: lakhs per annum, yearly pay in Indian rupees (1 lakh = 100,000). See [can I ask what narrowed the list?](https://guide.kello.ai/kello/your-list.md#can-i-ask-what-narrowed-the-list)

Kello also tells you plainly when a brief clashes with the market, for example when the skill you want usually sits in a different kind of company from the ones you named.

**Real example: the same role, briefed two ways.** A finance role first went in as a sparse brief: a few bullets, no target companies. Its lists came back with 4, then 2, then 3 candidates. The rewritten brief named target banks, NBFCs (Non-banking financial company: an Indian lender that isn't a bank.) and fintechs, gave a pay sweet spot ("currently drawing 50 to 65 lakhs") and spelled out exclusions ("no investment bankers, no consulting backgrounds"). It returned 15 candidates on the first pass.

## Should I name companies to target or avoid?

It helps. Name companies whose people you'd like to meet, and companies you don't want to hire from, such as customers or partners.

Kello already starts with your company and the kind of people it has hired. Your list adds what Kello can't see from the outside.

With each list, Kello reports on the companies you named: either the list includes people from them, or nobody from them cleared this round. When nobody cleared, Kello offers a dedicated pass on just those companies. Take it.

**Real example: the dedicated pass.** For a defence-sector sales role, nobody from the named companies made the first list. The recruiter asked for the dedicated pass. It returned 7 candidates, mostly from one of the named companies.

## How do I rule out the wrong segments?

Name them. One line of exclusions removes whole groups that would otherwise fill your list:

```text Example exclusions wrap theme={null}
No services companies, no semiconductors, no very traditional companies.
No marketplaces. Look at FMCG, e-commerce and D2C brands instead.
No investment bankers, no consulting backgrounds.
```

Kello confirms exclusions back to you and carries them through every list.

## How do I say what I mean by depth?

Draw the line when a title could mean several things. Without it, "senior ML" or "AI engineer" can bring you the wrong kind of good candidate.

```text Example depth lines wrap theme={null}
Strong core ML depth, not someone who has mainly built apps on LLM APIs.
Not looking for high ownership; we'd like to avoid product engineers.
Must be on a bug bounty leaderboard or high in CTF rankings.
```

## Should I share the context behind the JD?

Yes. Why past hires or applicants didn't work out often says more than the JD.

**Real example:** One recruiter forwarded a post about why direct applicants hadn't worked. Kello replied that it was *"prioritizing technical depth over 'flashy' profiles"*, and looked for active code contributions and technical writing.

A line is enough:

```text Example context line wrap theme={null}
Our last two hires came from large companies and struggled without structure.
```

## What makes an addition strong or weak?

Specific lines give Kello something to read profiles against. Vague ones don't.

| Weak | Strong |
| - | - |
| "Competitive salary" | "Up to 40 LPA fixed" or "No fixed budget, use market rates" |
| "Budget is flexible" | "25–35 LPA, can stretch to 40 for the right person" |
| "Good culture fit" | "Has worked in a team of under 20 engineers" |
| "Strong sales background" | "Has sold B2B SaaS to mid-market finance teams" |
| "Senior ML engineer" | "Core ML: has done post-training, not only apps on LLM APIs" |
| "Bangalore preferred" | "Bangalore, 3 days a week in office. Firm." |
| "Top companies" | "Would love people from fintech startups. Avoid our customers: `[names]`" |
| "Fast-paced team" | "Has shipped weekly in a startup of under 50 people" |

Kello never considers age, gender, religion, race, marital status or similar personal traits. If a brief asks for one, Kello leaves it out rather than swapping in a stand-in, such as an experience cap for an age. See [what if my brief asks for a personal trait?](https://guide.kello.ai/kello/trust.md#what-if-my-brief-asks-for-a-personal-trait)

## How do I share example profiles?

Sharing a few example profiles is the fastest way to show Kello your team's taste before the first list. Paste the LinkedIn links into your brief, each with a line on why that person is the template.

```text Example profiles in a brief wrap theme={null}
People I'd hire in a heartbeat:
https://www.linkedin.com/in/...  (strong frontend engineer who also joins sales calls)
https://www.linkedin.com/in/...  (built our payments stack at a 20-person startup)
```

The why matters. A bare link tells Kello who; the line tells it what to carry through the search. When one founder explained that their example was "not only a good frontend engineer" but someone who could "jump on sales calls", Kello looked for that trait across the whole list.

A CV works as an example too. Attach it and say `more people like this`.

You can also add examples mid-search. Reply with a profile, or name a person from your list, and Kello builds the next list around them.

### What does Kello learn from them?

Kello reads each example the same way it reads every candidate: for meaning, not keywords. It looks at what each of their employers actually does, the size and stage of those companies when they joined and when they left, how quickly they were promoted, and what they worked on.

Then it looks for what the examples share. Maybe they all joined companies early and stayed through a growth phase. Maybe they all moved from engineering into customer-facing work. Those patterns are often things a JD never says. Kello then finds more people like them.

### Who makes a good example?

* People on your team who are great at this job.
* Past colleagues you'd hire again.
* People at other companies you admire but can't hire.

If two quite different people would both succeed in the role, share both. Kello looks for several different kinds of people who could succeed, so the list isn't one profile repeated 20 times.

### Will the example people appear in my list?

No. They're examples, not candidates. Kello uses them to calibrate the search, and they won't be in your list.

If you're sharing people already in your pipeline (People you're already talking to or interviewing for this role.), say so. Kello treats them as a quality bar and won't contact them.

### What if I don't have any examples?

That's fine. Example profiles are optional. Kello always starts with your company and the kind of people it has hired, which gives the first list a strong starting point.

### Do my Yes marks work the same way?

Yes. Once your list arrives, every person you mark **Yes** or **No** teaches Kello the same way your examples did. Kello studies the Yes and No profiles in depth to learn your team's taste.

A short note on why (`love this one`, `too senior`) makes the reason explicit, so Kello doesn't have to guess. See [do I need to add a note?](https://guide.kello.ai/kello/feedback.md#do-i-need-to-add-a-note)

## Can I send several roles in one email?

No. Send **one role per email**. Every role gets its own search and its own list.

## What if I get something wrong?

Reply with the correction, like `it's a Director role` or `Pune, not Bangalore`. Kello restarts on the corrected brief, even mid-search.

Kello also confirms how it read the role within minutes of your email. That's the easiest moment to catch a misread.

## Related

> [card] [Brief templates](https://guide.kello.ai/kello/brief-templates.md)
>
> A general template, plus copy-paste emails for seven roles.

> [card] [How Kello evaluates](https://guide.kello.ai/kello/how-kello-evaluates.md)
>
> How Kello reads a profile.

> [card] [Your list](https://guide.kello.ai/kello/your-list.md)
>
> What arrives, and what happens when few people fit.

> [card] [Feedback and next lists](https://guide.kello.ai/kello/feedback.md)
>
> How your marks and notes shape the next list.

Send your next role to [source@kello.ai](mailto:source@kello.ai).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.