EmpCloud — Delivery intelligence

Ranks measured GitHub + Hubstaff signal only — not a performance review, not developer worth.

Generated by an AI model reading commit/PR content, complexity and tracked time. It cannot see DevOps, incident response, mentoring, architecture, code review depth, or work outside the connected repos — read each person's caveats before drawing any conclusion about them.

Conclusion

claude-opus-5 · period Fri Aug 28 2026 00:00:00 GMT+0000 (Coordinated Universal Time) – Sun Sep 27 2026 00:00:00 GMT+0000 (Coordinated Universal Time) · generated 2026-09-27 08:34 UTC · triggered by systemd:ranking · 23/25 people had evidence to rank on · ~$0.805

**The period was heavy on shipping and very light on review.** Nobody opened or reviewed a single PR, so every ranking below rests on commit titles, infra events and tracked time. That is a real gap in the org's evidence, not just in this report.

**Infrastructure carried a lot of the load.** Pankaj Kumar and Harsh Patel both paired substantive application commits with sustained AWS work across 19 and 17 distinct days. Rushendra Varma Mudunuru wrote no code at all but ran account and access work on 20 separate days, which is the most consistent ops record here.

**Volume and value did not always line up.** Several people are far above the median 33 commits, but some of that is repository scaffolding, backup dumps and "initial commit" work rather than considered change. At the other end, a few people logged full months of hours with no commits and no mapped AWS identity, so we simply cannot see what they did. Those cases need a manager conversation, not a low score.

Ranked (GitHub + Hubstaff evidence, evaluated by Claude)

Rank reflects measured delivery signal in this dataset, not overall developer worth — read the caveats.

#	Name	Tier	Confidence	Rationale	Caveats
1	Pankaj Kumar	top	0.55	Pankaj did real provisioning, not just deploy churn: his AWS trail shows CreateApp and CreateBranch on Amplify, CreateHostedZone on Route53, and CreateRole plus AttachRolePolicy on IAM, spread over 19 distinct active days, the second highest in the team. On the code side his 71 commits lean toward integration work with clear titles, like "Add isolated TheSportsDB proxy under /api/sportsdb" and "Enhance article and series pages with CMS integration and new sports", and 35 of them are refactor rather than chore. That combination of standing up new hosting and wiring the CMS behind it is the strongest single case this period. Brazil_CMS@a20ec6c cms@26be58b ecuadorfocus.com@f3fa398 puntoecuador.world@1bce1ac	1521 infra events is by far the highest number here, and a lot of it is repeated Amplify StartJob deploy traffic, so read the 19 active days rather than the raw count. No PRs opened or reviewed, so there is no independent sign anyone else looked at this code. Commit titles like "Initial commit from Create Next App" are scaffolding, not authored work. No connected task tracker confirms these changes were what was asked for. Not visible here: any production firefighting he did by reading logs, plus architecture calls on how the CMS proxy should be shaped.
2	Harsh Patel	top	0.40	Harsh ran deployments across 17 active AWS days and 19 distinct services, with Elastic Beanstalk environment updates, CloudFormation stack updates and load balancer target register and deregister calls, which is what a real release cycle looks like. His 99 commits back that up with product work that reads as considered: "Add Nexus Service external integration API with live progress polling", "Trend ingestion Phases 2-4, Apify sources, content briefs, scale tuning" and an "AdminAuditLog + four-eyes approval" change that is a genuine control, not a cosmetic one. He also shipped EN and PT-BR translation across the Nexus frontend. NexusAI@009760f NexusAI@0432529 coxa-hub@b88d0dc NexusAI@0bd983e NexusAI@7b527dd	1072 infra events is inflated: each Beanstalk deploy fires many autoscaling and load balancer calls on its own, so the 17 active days is the honest figure. 42 of 99 commits fall into "other", and titles like "pod setup and gpu changes draft" are vague about what actually landed. No PRs opened or reviewed, so no review trail exists for any of it. All infra work sits in the 1touch Cloud Infrastructure account. Not visible here: the debugging behind the "facefusion + heygen fixes" work and any GPU or vendor troubleshooting that left no AWS write.
3	Rushendra Varma Mudunuru	top	0.40	Rushendra has zero commits and still belongs near the top, because his AWS trail is the most consistent ops record on the team: 285 write events across 20 distinct active days and 17 services, the highest active-day count here. The work is account and access management rather than deploy noise, including CreateAccessKey and TagUser on IAM, repeated ModifySecurityGroupRules, RequestServiceQuotaIncrease and AddCommunicationToCase on an AWS support case. Raising quotas and chasing support tickets is exactly the kind of unblocking work that never shows up as code.	No commits and no PRs at all, so nothing can be said about his code quality this period. Tracked activity percentage is 10.8, the lowest of anyone with a full month of hours, and 153.4 hours all sit on an unmapped Hubstaff project. Repeated ModifySecurityGroupRules on the same security group can be trial and error rather than distinct changes, so the event count overstates the number of decisions made. Not visible here: the support-case back and forth itself, capacity planning conversations, and any access requests he handled by hand or over chat.
4	Aziz Bohra	top	0.40	Aziz worked both sides of the email-service build. In code, titles like "integrate auth readiness controls", "prove mailbox recovery reconciliation" and "align DEV1 auth readiness with frozen contracts" read as careful work against an agreed contract, and 21 of his 178 commits are tests, which almost nobody else here has. In AWS he made the matching domain setup real, with CreateEmailIdentity and PutEmailIdentityDkimAttributes on SES alongside Route53 record changes, across 12 active days. email-service@63ffd78 email-service@9b23c87 email-service@032fcda email-service@a620fcc	178 commits is more than five times the team median of 33, which raises the question of whether single changes were split across many commits. The titles read substantively, so this is a flag to check rather than a finding. "add single-EC2 homepage deploy" at 8139 churn is a big change with no review on it, and no PRs were opened at all. All infra work is in the 1touch Cloud Infrastructure account only. Not visible here: the DNS and deliverability debugging that happens outside AWS, such as checking propagation or working with a registrar.
5	Anshuman Parmar	top	0.55	Anshuman built the delivery intelligence tooling itself, with "Delivery intelligence platform: ingestion, scoring, and evidence UI" and "Add Notion as a work source, plus a Tasks page and per-person flags", and also did the POSITION data groundwork in "Complete Phase 0 data registry, schemas, and S3 load/validate/version". His AWS work over 11 active days and 12 services is deploy and cleanup: Beanstalk UpdateEnvironment and CreateApplicationVersion, plus DeleteStack, DeleteSecurityGroup and DeleteLaunchTemplate, which is tidying old environments away rather than just pushing new ones. Hubstaff-GitHub-Monitoring@61e805f Hubstaff-GitHub-Monitoring@d575cd6 POSITION_INTELLIGENCE@d33cc15 Hubstaff-GitHub-Monitoring@13a6998	31 of 75 commits are categorised "other", and "Added Dashboard" at 45052 churn is a very large change behind a two-word title. 280 of his 302 infra events sit in "account 1" rather than the main cloud infrastructure account, so most of that activity is one project's context. No PRs opened or reviewed this period. Not visible here: architecture decisions about how the scoring and evidence model should work, which is the kind of judgement this product depends on and leaves no commit.
6	Ankur Kumar	solid	0.55	Ankur covered three products and moved Nexus storage onto S3, which is a real platform change rather than a surface one. Other strong titles include "Add internship platform: apply flow, university partnership, newsletter" and an admin endpoint to hide demo and test data, which is the sort of small, well-scoped fix that keeps a demo usable. His AWS activity spans 13 active days, above the team median of 2. NexusAI@db85c53 coxa_crm@4d22e8b coxa-hub@a6f6246 NexusAI@4b0162b	The infra sample is almost entirely ModifySecurityGroupRules on the same two security groups, repeated over several days. That is low-variety work and the 96 event count overstates it. Two commits with identical titles and identical 2616 churn on POSITION campaign integration suggest a duplicate or re-push rather than two pieces of work. 102 commits with 22 in "other" and a "Initial commit from Create Next App" scaffold means the headline count flatters the picture. Not visible here: onboarding or support for the Coxa internship users, and any coordination across the three products he touched.
7	Abhishek Kulkarni	solid	0.55	Abhishek's 113 commits carry the largest feature work in POSITION_INTELLIGENCE and ZeroHumanMultiAgentV3, including "implement complete autonomous agentic system", "add provider-neutral deployment manager" and "add first-class Claude/Anthropic execution adapter". The bug fix "close finished mixed Ask jobs and stop stale Need" reads like someone who actually chased a queue problem down. He also stood up a database backup bucket and launched an EC2 instance for the Coritiba work. POSITION_INTELLIGENCE@2ff2056 ZeroHumanMultiAgentV3@2a540a3 ZeroHumanMultiAgentV3@e8f2c50 PulseDesk-Test@35f4be9	Most of his 54 infra events are ConsoleLogin and CheckMfa sign-in records, not changes. The genuine infra here is a handful of calls: CreateBucket, CreateKeyPair and RunInstances across 6 active days. Single commits at 8882 and 5638 churn are very large drops of code with no review on them, which makes them hard to check. "establish PulseDesk engineering foundation" is the only commit carrying a ticket reference, so most work cannot be traced to a request. Not visible here: design work behind the agentic system, and any unmerged branch work that did not land.
8	Jitu Prosad Saha	solid	0.40	Jitu sits right on the median at 33 commits but the titles are among the clearest here, and they are operational rather than cosmetic: "add SES bounce complaint suppression handling", "add inbound pipeline reconciliation and evidence harness", "add production backups and monitoring skeleton" and "harden manager mailbox backend contracts". Bounce suppression and backups are the unglamorous work that keeps an email product from getting blocked. His AWS work matches, with SES identity creation and Route53 records, plus setting up his own MFA device. email-service@68cfba8 email-service@a3f357b email-service@9dcaa85 email-service@23c8f0c	Two commits share the title "add production backups and monitoring skeleton" with nearly identical churn of 2286 and 2282, which looks like a re-push rather than two pieces of work. Only 13 infra events across 2 active days, which is at the team median and mostly one day of DNS and SES setup. No PRs, so none of this backend hardening was reviewed by anyone. Not visible here: deliverability monitoring and any mailbox issues he handled by hand after the code landed.
9	Rushikesh Powar	solid	0.40	Rushikesh handled the CMS-to-site integration across a wide set of properties, with "Integrating 6 websites vozdelecuador.com, informeec.com, conexionec.c" and a network-wide "Ship network SEO remediation for all 25 public sites (phases 0-5)" that is scoped and phased rather than ad hoc. He also removed the legacy PostgreSQL collection runtime in POSITION_INTELLIGENCE, which is deletion work most people avoid. On AWS he created and encrypted an S3 bucket for incident collection and kept an EC2 instance running across 8 active days. Global_CMS@0409205 cms@0a3f33d POSITION_INTELLIGENCE@6a4ca9c EcuadorCentral.com@a3880bb	Repeated StartInstances on the same instance across several days is routine, low-value activity, and 25 events is a small footprint overall. "update" at 8055 churn on EcuadorCentral.com is a placeholder message on a very large change, which makes it impossible to tell what shipped. 16 of 52 commits are "other" and no PRs were opened. Not visible here: the manual checks across 25 live sites that SEO remediation normally needs.
10	MOHD NAVED	solid	0.40	Naved has only 19 commits, well below the median of 33, but they are among the best-written here and each one describes a real problem solved: "queue offline writes, add Club/FanBox/POS PWAs, and fall back", "survive API timeouts and brief offline without fake writes" and "resolve QA fan portal bugs BF-1, BF-2, and BF-3" with actual bug references. He also did performance work, cutting re-renders and duplicate list API calls, and shipped DSR cross-system deletion, which is a compliance requirement. This is low volume and high care. coxa-1touch@cf96cab coxa-1touch@ff64930 coxa-1touch@ccfaa71 coxa-1touch@c8b12b4	His 3 infra events are only his own MFA setup and password change, so there is no infrastructure contribution to read here either way. 19 commits against 191 tracked hours is a thin ratio, and the hours are almost all on an unmapped project, so we cannot see where most of the month went. No PRs opened or reviewed. Not visible here: the QA loop behind the BF bug fixes, and any offline or device testing that produced no commit.
11	Dhiraj Ambekar	solid	0.55	Dhiraj's 34 commits sit at the median but the mix is healthy, with 10 refactors and product work that names its user need, such as "bulk edit UI for personas, bots, and campaigns" and "filter groups by persona gender and age range". He also fixed a real post-merge break in the Apify scraper. His AWS work is modest but spread over 8 active days and includes a Beanstalk environment update with the matching CloudFormation stack update, plus Route53 record changes. Social_Media_Automations@8392093 Social_Media_Automations@8db9328 Social_Media_Automations@5650b07 Social_Media_Automations@4e5c0bd	"update the current code of dummy number" at 1482 churn is an unclear title on a sizeable change. Several commits describe polish and restyling, which is useful but lower complexity than the platform work higher in this list. 32 infra events is near the low end and includes sign-in records. Not visible here: demo preparation and the customer-facing feedback loop that clearly drove the UI polish work.
12	Manish Chandrakar	solid	0.40	Manish ran the widest spread of any developer here, touching ten or more global CMS repos with real per-market work like "Add Italy sports hub, harden Serie A APIs, and soft-fail flaky upstream" and "Retarget HarimauGoal to Liga Super Malaysia and drop World Cup". Hardening an API and soft-failing a flaky upstream is a proper reliability change, not a cosmetic one, and he added pre-commit lint and build checks to at least one site. global-cms-Italiano@dbe84da global-cms-Malay@d7e9bb8 ecuadorenvivo.com@1b70e1a global-cms-Korean@4224366	122 commits with 51 in "other" and titles like "first commit" and "husky test" means a chunk of the count is scaffolding and setup, not authored change. The headline number is well above the median of 33 for that reason. A lot of the work is rebranding and palette changes repeated across sites, which is real but repetitive. infra.totalEvents is 0, which here most likely means no AWS identity is mapped to him rather than that he did no infra work. Do not read it as a gap. Not visible here: any hosting or Amplify setup he did under a shared account, and the content or translation checks each market site needs.
13	Akash Kumar	solid	0.55	Akash's real contribution is the email-service front end and delivery controls, with "complete day 6 delivery controls", "billing usage dashboard and gated ops screens" and a specific regression test for a stale host validator rejecting image keys, which shows he went back and pinned a bug down. On AWS he did the domain onboarding work, creating SES email identities with DKIM attributes and the matching Route53 records across 4 active days. email-service@7978ad7 email-service@74c0116 email-service@445a2c8 server-codebase-backup@0957d77 server-codebase-backup@df0b34d	Two commits in server-codebase-backup account for roughly 950,000 churn between them. These are bulk code dumps, not authored work, and they badly distort any volume reading of this person. Several commits are labelled by day number rather than by what changed, such as "Day 3" and "Day 4 cleanup", which tells a reader little. No PRs opened or reviewed. Not visible here: the mailbox and DNS verification steps done through vendor consoles outside AWS.
14	Pramod Soni	solid	0.40	Pramod delivered consistent per-language sports sections across the global CMS estate, including a Vietnamese football build, a Russian football hub via the CMS proxy, and cricket sections for Punjabi and Tamil with fixtures, teams and match detail. His 30 commits are just under the median but 9 are refactors and the titles are specific about what each site gained, which is better than the raw count suggests. global-cms-veitanamese-new@a993ee5 global-cms-Russian@34676db global-cms-punjabi@ecb41f1 global-cms-turkish-@253bb29	The work is the same pattern repeated across many language sites, so complexity per change is moderate. One commit is mostly an .env.example and gitignore update, which is setup rather than product work. infra.totalEvents is 0, which most likely means no AWS identity is mapped to him, not that he avoided infra work. Not visible here: language and content review with whoever checks these market sites before they go live.
15	Rishav Raj	solid	0.40	Rishav made only 6 commits, far under the median of 33, but every one is a feature and several are substantial: "complete migration phases" at 14577 churn, "add content aggregation intelligence workflow", "add Google Search scraper support" and a complete Custom Scrapers frontend workspace. This is the profile of someone doing a small number of large things rather than padding a count, so the low number is not a problem on its own. POSITION_INTELLIGENCE@a78df0d POSITION_INTELLIGENCE@09edede Social_Media_Automations@6879d25 POSITION_INTELLIGENCE@95015d4	6 commits against 190.8 tracked hours is the widest gap between time and visible output on the team. That is worth a conversation, not a conclusion. "complete migration phases" is a vague title for the largest change he made, so a reader cannot tell what migrated. infra.totalEvents is 0, which may simply mean no AWS identity is mapped to him. Not visible here: scraper tuning and data quality checking, which is largely run-and-inspect work that leaves no commit, plus anything sitting on an unmerged branch.
16	Pranav Ojha	solid	0.40	Pranav stood up four new Ecuador news sites and built a Dev Console for intelligence ops, and did follow-up quality work with "Translation correction, layout correction, SEO enhancement" on the Hindi site and "correction based on audit report" on the Indonesian one, which shows he acted on review findings. The problem is readability: 32 of his 35 commits land in "other" and four of the ten largest are just "initial commit". lavozec.com@2c1c466 POSITION_INTELLIGENCE@0ee7869 global-cms-hindi@fba9fb9 globla-cms-Indonesian@a2d752b	Four separate commits titled "initial commit" at 11,000 to 12,500 churn each are site scaffolds. High churn here reflects generated project files, not authored complexity. Commit messages are consistently thin, which makes it hard for anyone to check what changed without opening the diff. His 3 infra events are only his own password and MFA changes, so there is no infrastructure signal either way. Not visible here: the audit work itself, and the editorial and translation judgement behind the correction commits.
17	Cursor Agent	needs-attention	0.07	This is an automation account, not a person, so treat the rank as bookkeeping. It made 6 small email-service commits with clear, narrow titles such as "Add manual domain DNS verification" and "Promote domain after representative mailbox create", all under 520 churn. The work is coherent and sits alongside the human email-service effort rather than standing on its own. email-service@b9bc59d email-service@048305b	Evidence confidence is 7 percent and there are no tracked hours, so this cannot be compared with a person's month. Credit for this work most likely belongs to whichever engineer drove the agent, which the data does not show. No infra events and no PRs. Not visible here: who prompted and reviewed these changes.
18	Claude Code	needs-attention	0.07	Another automation account with a single 44-line commit, "Fix AttributeError crashing entire Gemini structuring batches". It is a well-titled fix that names both the error and its blast radius, but one small commit is all the evidence there is. Football_Scraping_Main@5f168e4	Evidence confidence is 7 percent, no tracked hours, no infra events, no PRs. Credit belongs to whoever ran the agent, which is not recorded. Not visible here: the debugging that found the crash in the first place.
19	mohammad saad	needs-attention	0.40	169.5 tracked hours on Thunder Marketing's Project, with no commits, no PRs and no AWS events. There is nothing to read here about what was delivered, which is a visibility problem first and a performance question only second.	infra.totalEvents is 0, which may simply mean no AWS identity is mapped to him. If his role is ops-leaning, this report is blind to all of it. All 169.5 hours sit on a Hubstaff project that is not mapped to any repository, so we cannot link time to output. A manager should confirm what the month was spent on before anyone draws a conclusion from this entry. Not visible here: design, research, QA, support or any work done outside GitHub and AWS.
20	Anmol Music	needs-attention	0.40	122.7 hours tracked, all on a project called Music, with no commits, no PRs and no AWS activity. The project name suggests work that is not software delivery at all, in which case this ranking is measuring the wrong thing for this person.	No code or infra surface exists for this person, so the dataset has nothing to evaluate. If the role is not an engineering one, they should be excluded from this report rather than ranked low in it. infra.totalEvents is 0, which here most likely means no AWS identity is mapped. Not visible here: creative or media production output of any kind.
21	Srija Bandapally	needs-attention	0.40	16.1 hours tracked on Thunder Marketing's Project and nothing else: no commits, no PRs, no AWS events. That is a small slice of a month and not enough to judge delivery either way.	There appear to be two records under this name, ids 17 and 179. They should be merged before the next period so the time is read as one person. 16.1 hours is far below the team median of 161.9, so this may be a partial period or partial tracking. infra.totalEvents is 0, which may mean no AWS identity is mapped rather than no infra work. Not visible here: anything done off GitHub, including analysis, meetings or documentation.
22	Shakir Khan	needs-attention	0.40	One hour of tracked time on the CMS project and nothing else in the period. There is a record, so he is ranked, but a single hour cannot support any read on delivery.	One hour against a team median of 161.9 suggests a new starter, a part-time arrangement or tracking that was not switched on. No commits, no PRs, no AWS events. Not visible here: onboarding, shadowing or any work done before tracking started.
23	Srija Bandapally	needs-attention	0.40	1.4 hours tracked on Strategy and Financial Consultant, with no commits, PRs or AWS activity. This looks like a duplicate or non-engineering record rather than a developer month.	Likely the same person as id 17. The two records should be merged or one retired. The project name suggests non-engineering work, which this report is not built to assess. No code or infra surface at all, so there is nothing to read. Not visible here: consulting or analysis output, which never touches GitHub or AWS.

Not ranked — no evidence this period

Listed, never guessed. Ranking someone here would fabricate a comparison the data can't support.

Cursor Agent, Ruhani Allagh

literally no GitHub, Hubstaff or AWS evidence this period

What this data cannot see

Code review and collaboration: nobody opened, merged or reviewed a single PR this period, so we cannot see who helped whom, who caught problems, or whose work was checked before it shipped.
Production incident response that leaves no AWS write: reading logs, restarting something by hand, a phone call or a Slack thread during an outage is invisible here.
Architecture and design decisions, which shape months of work but produce no commit and no CloudTrail event.
Mentoring, onboarding and unblocking other people, which several of the higher-ranked engineers are almost certainly doing.
Work sitting on unmerged branches, in local environments, or in repositories not connected to this dataset.
AWS activity for anyone without a mapped identity. A zero infra count often means the mapping is missing, not that no infrastructure work happened.
Hubstaff project mapping is mostly unconfigured, so tracked time usually cannot be linked to the repository it funded. That is our setup, not anyone's focus.
Non-engineering contributions such as QA, content, translation, research and customer conversations, which this report has no surface for.