आवेदक
How Aavedak works
Where the jobs come from, how referral emails are sent, and how the ATS engines score your resume. Aavedak recommends and prepares. You decide and send.
Overview
The flow
1
Discover
Fresh junior roles from company career pages, ranked against your resume.
2
Prepare
Check your resume with ATS engines and keep resumes, cover letters and templates together.
3
Apply and track
Every application sits in a tracker with its status history and follow-ups.
4
Ask for referrals
Send referral and follow-up emails from your own Gmail, only after you confirm.
You sign in with Google. New accounts are approved by an admin before the workspace opens.
Jobs
How Discover jobs works
Every night Aavedak checks the career pages of 804 companies, keeps only the junior engineering and tech roles that fit, and recommends the best matches to you. It never invents jobs, and no job is shown without a link to the original posting.
Where the jobs come from
Job-board feeds behind career pages
Most companies' career pages (careers.company.com, “Join us” pages) are powered by a hiring platform: Greenhouse, Lever, Ashby, SmartRecruiters or Workable. Each platform publishes the company's open roles through an official public API. Aavedak reads that feed, which holds the same postings the career page shows, and links you to the posting itself.
Companies' own career pages
Yes, dedicated career pages are supported too, when the page publishes standard job data (schema.org JobPosting, the format Google Jobs reads). Aavedak reads only the page it is given, respects the site's robots.txt and doesn't crawl the rest of the site. Admins can add any such career page.
Not covered yet: career pages on platforms without a public job feed (such as Workday, Darwinbox, Oracle or SuccessFactors) that also don't publish job data on the page. LinkedIn, Naukri, Indeed, Glassdoor, Wellfound and similar job sites don't allow automated reading, so Aavedak doesn't scrape them. Found a role there? Add it yourself with Add job or Paste JD on the Jobs page.
Which companies are scanned (804)
These companies' job boards had current openings in India or open to India when they were checked, or were added by an admin. Each name links to the careers board Aavedak reads. Pages added since the last update may not be listed yet.
Greenhouse365
- 6sense
- a16z
- AB InBev
- Abnormal Security
- Acquia
- Adjust
- Adyen
- Affirm
- Affle
- Agoda
- Air Liquide
- Airbnb
- Airship
- Airtable
- Akoya
- Akuna Capital
- Algolia
- Alloy
- Alpaca
- AlphaSense
- Altruist
- Amwell
- Anaplan
- Anthropic
- Apollo.io
- Appian
- AppLovin
- AppsFlyer
- Aptos Labs
- Arize AI
- Asana
- AssemblyAI
- Astera Labs
- Attentive
- Automattic
- Axiom
- Backblaze
- Bandwidth
- Base
- BetterHelp
- Betterment
- Bill.com
- Bishop Fox
- BitGo
- Bitpanda
- Bitwarden
- Blend
- Block
- Bloomreach
- Bluestone
- Box
- Branch International
- Branch Metrics
- Braze
- Brex
- Bugcrowd
- Buildkite
- Bybit
- Calendly
- Calm
- Camp K12
- Canonical
- Carbon Health
- Careem
- Carrot Fertility
- Carta
- CB Insights
- Chainguard
- Charles
- Chime
- CircleCI
- Clear
- Cloudbeds
- Cloudflare
- CloudSEK
- Clover Health
- Coalition
- Cockroach Labs
- Cohere Health
- Coinbase
- Collibra
- Comet
- Commvault
- ComplyAdvantage
- Consensys
- Constructor
- Contentful
- Contentstack
- CoreWeave
- Cortex
- Couchbase
- Coursera
- Cresta
- Cribl
- Culture Amp
- Current
- Customer.io
- Databricks
- DataCamp
- Datadog
- Dataiku
- Degreed
- Descope
- Descript
- DevRev
- Dialpad
- DiDi
- Discord
- Doma
- Domino Data Lab
- DoorDash
- Doximity
- DriveWealth
- Dropbox
- Druva
- Duolingo
- Dyno Therapeutics
- Earnin
- Elastic
- Elation Health
- Endor Labs
- Equal Experts
- Expel
- Faire
- Fastly
- Feedzai
- Figma
- Figure
- Fingerprint
- Fireblocks
- Five9
- Fivetran
- Flatiron Health
- Flex
- Flexport
- Flow Traders
- Forbes Marshall
- Forter
- Forward
- Freenome
- Future
- Galileo
- Garner Health
- Gemini
- GetYourGuide
- Ghost
- Ginkgo Bioworks
- GitLab
- Glance
- Gleanwork
- Globalization Partners
- GoCardless
- Grafana Labs
- Graphcore
- Greenhouse
- Groww
- Gusto
- HackerRank
- Harness
- Headout
- Healthie
- Help Scout
- HeyGen
- Highnote
- HighRadius
- Honeycomb
- Honor
- HP
- Huntress
- IMC Trading
- Immunefi
- Inceptive
- India Quotient
- InfoSum
- InMobi
- Insider
- Instacart
- Intercom
- Invisible Technologies
- Island
- Isomorphic Labs
- Jane Street
- JFrog
- Jump Trading
- Justworks
- Karat
- Karbon Card
- Khan Academy
- Kite
- Klaviyo
- Knock
- Komodo Health
- Labelbox
- Lattice
- LaunchDarkly
- LayerZero Labs
- Leapwork
- Liftoff
- Lightmatter
- Lightning AI
- Lithic
- Lookout
- Lyft
- mabl
- Mattermost
- Maven Clinic
- Medium
- Melio
- Mercury
- Midi Health
- Mighty Networks
- Mixpanel
- Modern Health
- Moloco
- Momentic
- MongoDB
- Monzo
- Motive
- Mozilla
- N26
- Nansen
- Natera
- Neo4j
- Netlify
- Netskope
- New Relic
- Newsela
- Nexus Venture Partners
- ngrok
- NICE
- Nothing
- Nuro
- Observe.AI
- Ocrolus
- Octopus Deploy
- Okta
- OKX
- One Medical
- OpenZeppelin
- Oportun
- Orca Security
- Oscar Health
- Otter.ai
- Oura
- Pacaso
- PagerDuty
- Pantheon
- Papa
- Park+
- PathAI
- Patronus AI
- Payoneer
- Peak XV
- Peloton
- Pendo
- Pie Insurance
- Pilot
- Pine Labs
- PlanetScale
- Postscript
- Proton
- Public
- PubMatic
- Pure Storage
- Qualia
- Qualtrics
- Remote
- Ripple
- Riskified
- Robinhood
- Roofstock
- Rubrik
- Salesloft
- Salt Security
- Sama
- Samsara
- Sauce Labs
- Scale AI
- SeekOut
- Sendbird
- Shein
- Sigma Computing
- Sigmoid
- SingleStore
- Skillsoft
- Slice
- SmartBear
- Smartsheet
- Snorkel AI
- SoFi
- Spot
- Stability AI
- StackBlitz
- STAGE
- Starburst
- Storyblok
- StreamNative
- Striim
- Stripe
- Suki
- Sumo Logic
- SumUp
- Sword Health
- Taboola
- Tailscale
- Talkspace
- Tanium
- Tastytrade
- Teachable
- Tenable
- Tenstorrent
- Testlio
- The Trade Desk
- Thoughtworks
- Thunes
- Tia
- Tide
- Tines
- Toast
- Together AI
- Torq
- Tower Research Capital
- TripActions
- Tripadvisor
- TrueLayer
- Turing
- Twilio
- Typeface
- Typeform
- Uber Freight
- Udemy
- Ultimate
- Upgrade
- Upstart
- Vast.ai
- Vectara
- Veracode
- Vercel
- VerSe Innovation
- Verve
- Vim
- Virtu
- Vonage
- Wasabi
- Waymo
- Webflow
- Wikimedia Foundation
- Wiz
- Workboard
- Workleap
- WorldQuant
- Wrike
- Xendit
- Yes Bank
- Yotpo
- Yugabyte
- Zenoti
- Zeta Global
- Zocdoc
- ZoomInfo
- Zscaler
- Zuora
- Zwift
Lever88
- 100ms
- 15Five
- 1inch
- 360Learning
- Airalo
- Aircall
- Aleph Alpha
- AllTrails
- Anchorage Digital
- AngelList
- Arcadia
- Articulate
- Binance
- Blue Star
- Bolt
- Brilliant
- Canvas Medical
- Capital Float
- CertiK
- Cloudinary
- Coupa
- CRED
- Crypto.com
- dLocal
- Dozee
- Educative
- Esper
- Fi Money
- Finch
- Findem
- Fly.io
- Fundrise
- Gem
- Glide
- Hevo
- Hevo Data
- Included Health
- Jito
- Jobvite
- Kavak
- Kinsta
- Lalamove
- Level AI
- Lodgify
- Lyra Health
- Matillion
- Meesho
- Metabase
- Mindtickle
- Moonpay
- Neon
- Ninja Van
- Nium
- Offchain Labs
- Outreach
- Oxylabs
- Palantir
- Paytm
- Pigment
- Pipedrive
- Pocket FM
- Porter
- Prosper
- Qonto
- Quantum Health
- Ribbon Health
- Ro
- Safe Security
- Secureframe
- SonarSource
- Sophos
- Spotify
- Superside
- Sysdig
- Teller
- Toptal
- Truv
- Twingate
- Veeva
- Vida
- WalkMe
- Wealthfront
- Zerion
- ZeroTier
- Zeta
- Zilliz
- Zoox
- Zus Health
Ashby317
- 0x
- 1Password
- Abacum
- Abridge
- Acorns
- Airbyte
- Airwallex
- Akasa
- Alchemy
- Amplitude
- Analog Devices
- Andela
- Anyscale
- Apify
- Applied Materials
- Arkose Labs
- Arta Finance
- Ashby
- Asimov
- Astra Microwave
- Astronomer
- Atlan
- Atlas Copco
- Atlys
- Atomic
- Atomic Invest
- Attio
- Ava Labs
- Babbel
- Baseten
- Benchling
- Bounce
- Brainly
- Branch Insurance
- Browserbase
- Bruno
- Bubble
- Buffer
- Bunny.net
- Bureau
- Cambly
- Capsule
- Castle
- Causal
- Cerebras
- Chai Discovery
- Change Healthcare
- Character AI
- Checkly
- Chili Piper
- Circle
- ClassDojo
- Clerk
- ClickHouse
- ClickUp
- Close
- CloudZero
- Codat
- Coder
- CodeRabbit
- Cognition
- Cohere
- Collective Health
- Column
- Commure
- Composio
- Compound
- Cosine
- Crusoe
- Cube
- Cube Software
- Cursor
- CynLr
- Dave
- Decagon
- Deepgram
- Deepnote
- Demandbase
- Depot
- Ditto
- Docebo
- Docker
- Doppler
- Dovetail
- Drata
- Dune
- Eigen Labs
- Eight Sleep
- ElevenLabs
- Elliptic
- Ema
- EngFlow
- Etched
- Exa
- Factory
- Faros AI
- Fathom
- Fiddler AI
- Fig
- Finout
- Firecrawl
- Fireworks AI
- FlutterFlow
- Freetrade
- Front
- FullStory
- G2
- Gainsight
- Gamma
- Generate Biomedicines
- GitBook
- Gorgias
- Graphite
- HackerOne
- Harmonic
- Harvey
- Health Gorilla
- Helius
- Heron Data
- Hex
- Hightouch
- Hinge Health
- Hopper
- Horizon3.ai
- HUMAN Security
- Ideogram
- Illumio
- Incident.io
- Infinitus
- Infisical
- InfluxData
- Insitro
- Interview Kickstart
- Iterable
- Jellyfish
- Jigsaw Academy
- Juicebox
- Juniper Square
- Kin Insurance
- Kinetic Green
- Kiwi.com
- Kong
- Kraken
- Kustomer
- Ladder
- Lago
- Lambda
- LangChain
- Lark
- Leap Finance
- Ledger
- Lemon.io
- Lemonade
- Levels
- Lightdash
- Lightspeed India
- Linear
- LiveKit
- LlamaIndex
- Lovable
- Magic Eden
- Material Security
- Materialize
- Matter Labs
- Maven Securities
- Maze
- Mercor
- Merge
- Mesh Payments
- Method Financial
- Meticulous
- Metriport
- Middesk
- Midjourney
- Mintlify
- Miro
- Modal
- Modern Treasury
- Mollie
- Mosaic
- MotherDuck
- Mux
- Mysten Labs
- Nabla
- Navi
- Neptune.ai
- NetBird
- Nord Security
- Notable
- Notion
- Nous Research
- Nubank
- Oak Street Health
- Omni
- OpenAI
- Openly
- OpenSea
- OpsLevel
- Orb
- Origin
- Oscilar
- Oso
- Outschool
- Owkin
- Oyster
- Paddle
- Paradigm
- Paradox
- Parity Technologies
- Patreon
- Paxos
- Permutive
- Perplexity
- Persona
- Phantom
- Pika
- Pilgrim
- Pinecone
- Plaid
- Plane
- Pleo
- Plural
- Poolside
- PostHog
- Prefect
- Preply
- QuickNode
- Rad AI
- Railway
- Ramp
- Range
- ReadMe
- Recharge
- Redis
- Reducto
- Reka
- Render
- Replit
- Resend
- Rho
- River
- Rogo
- RunPod
- Runway
- Rutter
- Sanity
- Sardine
- Sarvam
- Scribd
- Semgrep
- Sentra
- Sentry
- Seon
- Sequoia One
- Sesame HR
- Sierra
- Sift
- SigNoz
- Singular
- Sitecore
- SiteMinder
- Snapdocs
- Snowflake
- Snyk
- Socket
- Socure
- Sona Comstar
- Sonder
- Speak
- Speakeasy
- Sprig
- Stable Money
- Steel
- Stigg
- Strava
- Stream
- Substack
- Succinct
- Supabase
- Surfshark
- SurveyMonkey
- Swarmia
- Synthesia
- Taktile
- Tavily
- Tekion
- Temporal
- Ternary
- Thinkific
- Tiger Analytics
- Tiger Data
- Titan
- Tonal
- Trading 212
- TRM Labs
- Turnkey
- Turquoise Health
- Twenty
- Unify
- Uniswap Labs
- Unit
- Unstructured
- Vanta
- Vantage
- Velocity
- Vivid Money
- Warp
- Watershed
- Weaviate
- Whoop
- WorkOS
- Writer
- Y Combinator
- Yotta
- Zapier
- Zed
- Zilch
SmartRecruiters34
Which jobs are kept
- Location
- In India, or remote and open to India (worldwide, anywhere, APAC or Asia). A plain “Remote” with no region is left out, because it usually means remote within another country.
- Every engineering role
- Any discipline: software, data engineering, analytics engineering, ML and AI, DevOps, cloud, SRE, QA and SDET, security, embedded and firmware, electronics, electrical, mechanical, civil, chemical, process and more. Indian junior titles that don't say “engineer” count too: Graduate Engineer Trainee (GET), Member of Technical Staff (MTS), Software Development Trainee, Technology Analyst, full-stack, frontend and backend developers.
- Other tech roles
- Data analyst and data scientist, business and BI analyst, machine learning and AI research, QA and testing, IT support and system, network and database administration, security analyst, UI/UX and product design, and associate product manager.
- Junior level
- The posting asks for under 3 years of experience. Senior, lead, staff, principal, architect and manager titles are skipped, as are internships.
- Not included
- Sales and pre-sales roles that borrow the word “engineer” (Sales Engineer, Solutions Engineer), engineering managers, and non-tech roles such as marketing, finance or recruiting.
How you get recommendations
- Each new job is scored against your resume and your career preferences: skills, preferred roles and locations. Jobs below the match threshold are not recommended.
- You get up to 50 new matches a day, best match first. They stay in Discover until you apply, bookmark or ignore them.
- Change your career preferences and Discover is re-ranked straight away. Don't want to wait for the nightly scan? Use Refresh jobs on the Jobs page.
- A job leaves Discover 30 days after it was posted, or as soon as the company takes it down. You can turn recommendations off in Profile settings → Job discovery.
Referrals
How the Referrals page works
The Referrals page is where you ask people at a company to refer you. It has three columns: your active applications, the email, and the people you're writing to. Mail is sent through your own Gmail, after you connect it with Google. Aavedak never sends anything you haven't confirmed.
1
Pick an application
Choose from your open applications. A job from Discover that you haven't applied to yet can be picked too; it's added to your tracker as Bookmarked first, because every referral email belongs to an application.
2
Pick people and a template
Select the people at that company you want to reach and an email template. Placeholders like {{company}} and {{role}} fill in from the application. You can attach one of your resumes.
3
Confirm, then 20 seconds to undo
Nothing is sent until you confirm. After that, a 20-second countdown runs on screen with an Undo button. Then the emails go out from your own Gmail account, one per person.
4
Follow up
The Already sent tab lists applications where you've reached out, with when each person was last contacted. Follow-ups use their own templates. To avoid repeated nudges, you wait at least an hour between follow-ups to someone who hasn't answered; once they reply, you can answer straight away.
Search and status filters at the top narrow the application list. The Needs referral tab shows applications you haven't reached out for yet. Email templates live on the Documents page.
ATS
How the ATS page works
Companies often screen resumes with applicant tracking systems (ATS) before a person reads them. The ATS page runs your resume through 11 different engines (1 native, 6 reference and 4 open source), so you can see where they agree and what is missing before you apply.
Three modes, picked from what you give it
Resume only
Just your resume. Engines score structure, wording, impact and formatting.
Role match
Resume plus a job title. Engines compare your resume with that role's usual skills.
Job match
Resume plus a job description. Engines compare your resume with that specific posting.
- Pick one or more resumes and engines. Each resume and engine pair runs on its own, and its progress shows live in the results table.
- Each score is shown as a ring with its verdict. Open one to see the breakdown: matched and missing skills, the checks behind the number, and suggested fixes.
- The Final score column is a weighted score of every engine that ran for that resume: Aavedak ATS counts for 50%, the reference engines 35% and the open-source engines 15%, shared out over whichever engines ran. Hover it to see each engine's weight.
- An engine that needs a job description or title you didn't provide shows Needs input instead of guessing.
ATS
The 11 ATS engines
Aavedak runs 11 engines: 1 native, 6 reference and 4 open source. Each one is tagged Native, Ref or OSS on the ATS page.
1NativeNative
Aavedak's own scorer. It credits skills only when your resume shows evidence for them.
Aavedak ATS
6ReferenceRef
Modelled on how popular resume checkers describe their scoring in public. They are approximations, not the vendors' own engines, and Aavedak is not affiliated with these companies.
Jobscan, Resume Worded, Teal, Rezi, SkillSyncer, Enhancv
4Open sourceOSS
Adapted from open-source resume analyzers, following the formulas in their source code. Each one links to its original project.
Open ATS, ATS Resume Checker, Resume Skills Extractor, Hybrid Resume Analyzer
- 01
Aavedak ATS
NativeNative evidence-based match & resume quality.
Aavedak’s best-of benchmark: 11 resume-quality criteria drawn from every engine (parseability, impact, verbs, bullets, wording, tense, spelling, dates/gaps, length vs years of experience, skills evidence) plus evidence-based skill, responsibility, title, experience and education match when a title or JD is given.
- Modes
- Resume only · Role match · Job match
- Inputs
- Job title optional · JD optional
- 02
Jobscan
RefJob-specific keyword & title match (reference).
Compares your resume to one JD. Match score ≈ hard skills 45% + soft 15% + other keywords 25% + title 15%. Readability is shown separately and does not change the match total. Needs a job description.
- Modes
- Job match
- Inputs
- Job title optional · JD required
- 03
Resume Worded
RefImpact, skills, wording & presentation (reference).
Share of Impact, Brevity and Style checks passed (equal weight, as Resume Worded describes), calibrated to your years of experience. With a JD, hard-skill targeting is blended in 30%.
- Modes
- Resume only · Job match
- Inputs
- Job title optional · JD optional
- 04
Teal
RefSeparate resume score vs job match (reference).
Two separate scores: Resume Score (share of Teal's documented Resume Analyzer issue cards cleared) without a JD, or Job Match (share of the JD's hard, soft and other keywords on the resume) with one.
- Modes
- Resume only · Job match
- Inputs
- Job title optional · JD optional
- 05
Rezi
RefContent, format & optimization readiness (reference).
Five readiness categories: content, format, optimization, best practices, and application readiness. With a JD, optimization reflects keyword targeting.
- Modes
- Resume only · Job match
- Inputs
- Job title optional · JD optional
- 06
SkillSyncer
RefWeighted hard/soft/title/degree match (reference).
Needs a job title and JD. Score out of 100: hard skills 60 + soft 15 + other keywords 5 + title 10 + degree 10. Keyword frequency matters; empty JD categories get full points. Title is all-or-nothing.
- Modes
- Job match
- Inputs
- Job title required · JD required
- 07
Enhancv
RefContent checks: parse rate, impact, repetition, spelling (reference).
Five Content checks from Enhancv's checker: ATS Parse Rate, Quantifying Impact, Repetition, Spelling & Grammar, Bullet length. With a JD, blended 70/30 with job tailoring (hard and soft skills, action verbs, title).
- Modes
- Resume only · Job match
- Inputs
- Job title optional · JD optional
- 08
Open ATS
OSSKeyword, formatting & content-quality scan (open source).
Open ATS weights from its source: keyword 50% (hard 50 / soft 25 / action verbs 15 / industry 10) + formatting 25% (penalties from 100) + content quality 25% (action verbs, passive voice, hedging, length). Title-only matches the role's skill profile; with neither title nor JD, keyword is skipped.
- Modes
- Resume only · Role match · Job match
- Inputs
- Job title optional · JD optional
- Source
- github.com/jlynshue/open-ats
- 09
ATS Resume Checker
OSS8 weighted categories with severity-ranked fixes (open source).
Keyword 22% + structure, formatting, writing, achievements 14% each + experience 10% + education, contact 6% each. Keywords come from the JD (or the title's skill list). Without either, keyword match is left out instead of counted as 100.
- Modes
- Resume only · Role match · Job match
- Inputs
- Job title optional · JD optional
- Source
- github.com/Jahangirhussen/ats-resume-checker
- 10
Resume Skills Extractor
OSSTF-IDF similarity + skill coverage by category (open source).
60% TF-IDF cosine similarity (1–3 word phrases, ×180 calibration, capped) + 40% share of JD skills found on the resume. Shows matched, missing and bonus skills by category, plus the raw similarity. With only a title, scores coverage of that role's skill profile.
- Modes
- Role match · Job match
- Inputs
- Job title optional · JD optional
- Source
- github.com/blueabstract/resume-skills-extractor
- 11
Hybrid Resume Analyzer
OSSSkill/keyword rubric combined with text similarity (open source).
Rubric from the reference (skills 40 + keyword overlap 30 + length 20 − 2 per missing skill, out of 90) normalized to 100, combined 70/30 with TF-IDF similarity. Both parts are shown separately. Title-only: skills + length out of 60. Resume-only: the source's rule-based resume validation.
- Modes
- Resume only · Role match · Job match
- Inputs
- Job title optional · JD optional
- Source
- github.com/Anirodh-Padhy/resume-analyzer
No engine here is the real screening system any particular company uses. Treat the scores as guidance: look for skills that keep showing up as missing across engines.
More about the project on the About page, and what changed recently in the changelog.