Introducing Brightspark: the tech jobs worth your time, ranked for you

Introducing Brightspark: the tech jobs worth your time, ranked for you

Brightspark is live. It collects tech jobs in Bengaluru and remote India straight from employers' career sites and open job boards, then ranks every one against your skills, with a plain reason for each score.

Today it tracks 71 employer career sites and 5 open job boards, holds about 1,068 open roles, and refreshes roughly every six hours. It is free, invite-only while we grow, and built so that nobody can pay to move a job up your list.

The problem: job hunting in Bengaluru tech is mostly grunt work

If you work in tech in Bengaluru, you already know the routine. You open five portals, scroll past the same role reposted by three recruiters, and click into listings that closed a month ago. The good roles are often sitting on a company's own careers page, which no aggregator bothered to check.

The pain points we kept hitting:

• Noise over signal. Portals rank by recency or by who paid, not by whether the role fits you.
• Stale and duplicate listings. The same opening shows up many times, and old ones never seem to die.
• Dealbreakers hidden in the fine print. Service bonds, night shifts, contract-to-hire and third-party payroll only surface after you have applied or interviewed.
• No salary context. Most postings say nothing about pay, so you go in blind.
• No guidance on gaps. A role asks for one skill you lack, and nothing tells you how to close it.
• Your data as the product. Many job sites earn money by selling your profile or letting employers buy visibility.

What Brightspark does about it

You tell Brightspark what you do once: your skills, the roles you want and your dealbreakers. You can paste or upload your CV to detect skills, then confirm them. From then on you open a ranked shortlist instead of a search box.

| Problem | What Brightspark does | | --- | --- | | Noise over signal | Scores every role on skills, seniority, work mode, salary and freshness, and shows exactly why. You can tune the weights, or let your shortlist and ratings suggest them. | | Stale and duplicate listings | Pulls from employers' own career sites and established boards, labels the source, shows real posting dates, and drops duplicates and anything older than 30 days. | | Hidden dealbreakers | Flags service bonds, night and weekend shifts, contract-to-hire, third-party payroll and travel requirements before you apply. | | No salary context | Shows a market estimate when the posting is silent, and a company-level estimate where data exists. | | Skill gaps | Suggests well-rated courses and the certifications that prove the skills a role asks for. | | Your data as the product | Your profile, notes and shortlist are visible only to you. Nothing is sold, and ads never change your scores. |

A few smaller touches matter more than they look: roles at your current employer can be hidden, next-step roles are suggested from your current title, and returnship programmes are called out for people coming back from a career break.

The tech stack: built to run cheaply

Brightspark is deliberately small. It runs on free job sources and no paid AI, so the cost of serving one more member stays close to zero. That is what lets it stay free.

How a job reaches your shortlist:

1. Collect. A scheduled collector reads the career sites we track and open boards such as Himalayas, about every six hours. It respects each site's rules for automated access and never scrapes LinkedIn or Naukri.
2. Clean. Listings are normalised, de-duplicated and tagged with their source and posting date. Anything older than 30 days is dropped.
3. Extract. Skills, seniority, work mode and dealbreakers such as service bonds or night shifts are pulled out of each posting.
4. Score. Each role is scored against your profile with transparent, weighted rules, so every number comes with its reasons.
5. Serve. You see a ranked list, with salary estimates and course suggestions alongside.

| Layer | What we use | | --- | --- | | Hosting and backups | DigitalOcean | | Backend, database and auth sessions | PocketBase | | Sign-in | Google sign-in (we hold no passwords) | | Frontend | \[framework\] | | Collector | \[language and scheduler\] | | Transactional email | SendPulse | | Salary estimates | OpenWeb Ninja (role, company, location and experience band only) | | Analytics and ads | Google Analytics and Google AdSense, only with consent |

Privacy shaped the design as much as cost did. The database is not reachable from the internet. An uploaded resume is read once in memory and discarded, never written to disk. Pasted resume text is parsed in your browser and never sent to us. The whole service is built to India's Digital Personal Data Protection Act, 2023.

The path to monetization: free first, never pay-to-rank

Brightspark is free, and the plan is to keep the core experience free. The model starts with one rule: money never touches the ranking. Nobody can pay to change your match scores or where a job appears in your list, and we never sell your data.

Today: ads cover the servers. Our main costs are servers and hosting. A few clearly labelled Google AdSense ads in the job list and on job pages pay for them. They appear only if you allow them, never in your shortlist, and you can choose personalised, non-personalised or no ads at all.

Next: keep costs flat as we grow. Invite-only batches let us add members without the bill outrunning the ad revenue. Free sources and rule-based scoring mean each new member costs almost nothing.

Later: options that fit the rules. As the member base grows, there are ways to earn more without breaking the promise above. Partner course and certification links in the skill-gap suggestions, clearly marked, are one. Optional premium extras for power users, such as instant alerts or deeper salary data, are another. Anything we add will be opt-in, labelled, and kept away from your scores.

What's next, and how to get in

We are opening Brightspark in small batches so it stays fast and the job list stays relevant. Next on the list: more employer career sites (members can already suggest companies), and more role families beyond DevOps, SRE, security, AI/ML and cloud.

Want in? Request an invite with your Gmail address, and we will email you a personal invite link as soon as there is room. Questions or feedback: brightsparkhq+hello@gmail.com.

One honest caution: we choose which career sites to track, but we don't verify individual listings. Always confirm a role on the employer's own site, and never pay a fee to apply.