Hi, I'm Pallavi. Technical Project Analyst.

I turn messy field processes into no-code apps, dashboards, and the delivery systems around them.
Proven across digital health and education, with AI built into the work, not just the conversation.

Portrait of Pallavi Verma

I take a process that lives in scattered forms or people's heads and turn it into a no-code app on platforms like CommCare, manage the rollout end-to-end, then build the dashboards that tell leadership whether it's working. Four years supporting project delivery across digital health and education.

Now: Open to project management, workflow implementation, and data analysis roles, across sectors. Self-teaching AI engineering through builds.

Get in touch  → See experience

By the numbers

The work, in numbers.

30+
No-code apps configured and deployed
on CommCare and Connect
15+
Dashboards designed and shipped
used by program leadership
~20%
Data inconsistencies reduced
via validation rules and case-linking
10+
Partners onboarded
onto the systems I supported
50+
Team members trained and onboarded
across implementation partners
4
Programs contributed to
delivered or supported, across India, Africa, and Tajikistan

What I focus on

Three things, done well.

01

Configuration & implementation

Configuring no-code apps and the workflows around them on CommCare and Connect: modules, role-based logic, conditional validations, case-linking, and payment-linked service delivery. From field discovery through partner coordination to a system that goes live and gets adopted. Built in digital health, the same approach fits any structured field process.

02

Data, dashboards & M&E

KPI dashboards in Excel and Power BI, data quality audits, drop-off and outcome analysis. MEL frameworks and CART principles for evidence that's Credible, Actionable, Responsible, and Transportable. Translating field reality into numbers leadership can act on.

03

AI & automation

End-to-end builds with Python, Flask, n8n on Docker, and the Gemini API. Self-taught, AI-assisted, shipped. See the job-application bot.

About

The short version.

I'm a Technical Project Analyst with 4+ years configuring no-code apps, the dashboards that read the data those apps collect, and the delivery systems around both. At Dimagi I worked across four digital health implementations on CommCare, three delivered end to end and one supported, including frontline-worker apps deployed through the Connect platform across India, Sub-Saharan Africa, and Tajikistan, and helped design the M&E systems behind them. Digital health and education are where the work happened, not the whole of what I do.

I bring an adaptive, builder mindset: I took on team leadership at Byju's, stepped into new Connect program areas as they emerged at Dimagi, and independently shipped a personal AI automation tool outside any formal training.

My engineering foundation (B.Tech, Electronics & Communication) is what makes the analyst side of this work feel natural, and what pulls me toward problems where program delivery, analytics, and technology overlap.

Toolkit

Tools I'm comfortable in.

A mix of platforms I've used at work and ones I picked up building personal projects. Per-role breakdowns live on the Experience and AI Job Bot pages.

Digital health platforms

CommCareConnect

CRM & support

SalesforceHubSpot

Project & program management

JiraAsanaConfluenceGoogle WorkspaceMicrosoft Office

Data, M&E, MEL frameworks

Excel (advanced)Power BISupersetCART principles

AI & automation (personal builds)

PythonFlaskn8nDocker Gemini APIPlaywrightBeautifulSoup WeasyPrintJinja2

Leadership & volunteer

A pattern of stepping up.

  • Dec 2024 to Present
    Vice-President, Alumni Society · Poornima University
  • Sep 2019 to Nov 2020
    Chair, Students' Council · Poornima University
  • Dec 2017 to Feb 2021
    Student Editor, Annual Magazine SPECTRUM

Get in touch

Open to project management, workflow implementation, and data analysis roles.

Across sectors. Available remote or for relocation to the right role. The fastest way to reach me is email.

✉

Email

Opens a Gmail compose window pre-filled with my address.

→
in

LinkedIn

For a fuller picture of my background and to connect.

→

Experience

Where I've shipped.

Technical Project Analyst @ Dimagi

May 2023 – Sept 2025 · Full-time · Remote (New Delhi, India)

  • Designed CommCare apps for three programs, covering modules, role-based logic, conditional validations, and case-linking. Translated field workflows into structured digital modules through discovery with clinical leads, frontline workers, and product teams.
  • Coordinated end-to-end project delivery across product, engineering, M&E, and field teams. Owned timelines, scope, partner readiness, and feature rollouts through Jira, Asana, and Confluence.
  • Contributed to M&E systems across 3 program implementations, including 15+ KPI dashboards I designed in Excel plus drop-off analysis and data-quality workflows used by program leadership. Drove a ~20% reduction in data inconsistencies through validation rules and case-linking I designed.
  • Trained and onboarded 100+ frontline workers across multiple implementation partners through live sessions and structured documentation (SOPs, user manuals, field-ready job aids). The cohorts I worked with sustained 90%+ platform adoption per internal reporting.
  • Repeatedly took on new Connect program areas as they emerged at Dimagi, an adaptive ownership style recognized by leadership.
  • Monitored global digital health program implementations across Tajikistan, Ethiopia, Uganda, Zambia, Kenya, and India, ensuring smooth rollouts tailored to partner and end-user needs.
  • Planned and executed launches with field teams, product leads, and engineers across multi-country rollouts. Sustained partner trust through structured onboarding, troubleshooting, and post-launch engagement.

Tools: CommCare · Connect · Jira · Asana · Confluence · Power BI · Excel · Google Workspace

Product Specialist @ BYJU'S (Think & Learn Pvt Ltd)

Aug 2021 – Jan 2023 · Full-time · On-site

  • Managed onboarding and direct support for 200+ users in an EdTech product environment, building rapport across a wide range of profiles.
  • Identified recurring issue patterns from user feedback and surfaced them to product and process owners for documentation and workflow improvements.
  • Took on team leadership at BYJU'S, leading 5 team members across onboarding workflows, issue resolution, and individual performance for service delivery to 200+ users.
  • Enabled data-driven decisions by tracking and reporting KPIs to leadership; delivered targeted performance feedback that improved retention.

Tools: Salesforce · HubSpot · Excel · Microsoft Office

Subject Matter Expert @ Chegg Inc.

Sep 2020 – Sep 2021 · Part-time

  • Provided expert-level explanations and structured answers for academic problems in electronics and communication, used by students globally.
  • Built a habit of clear, audience-aware technical writing that I still rely on for SOPs and user manuals today.

Patent Research Analyst @ 3AIP Services Pvt. Ltd.

Feb 2021 – Apr 2021 · Internship · Chandigarh, India

  • Conducted patent landscape research and prior-art analysis across electronics and IoT domains.
  • Documented findings into structured timelines and reports for client stakeholders.

Projects

Three projects, one playbook.

Each built on CommCare via Connect at Dimagi. The patterns travel: discovery, workflow configuration, the M&E system that reads the data, partner enablement, and rapid iteration with field input.

Maternal & newborn health CommCare on Connect Delivered

Kangaroo Mother Care (KMC)

A CommCare workflow and M&E system for preterm and low-birth-weight newborn care, deployed across partners in India and Sub-Saharan Africa.

GeographyIndia, Sub-Saharan Africa
PartnersMultiple implementation partners
UsersFrontline workers and clinical leads
Program reach675+ newborns enrolled (Dimagi, public)

What it does

A CommCare app deployed on the Connect platform that lets frontline workers capture KMC care delivery in real time, track risk flags, and follow up on infants discharged from facility care. M&E layers feed real-time dashboards that partner program leads use to spot adherence gaps before they become outcomes problems.

My role

  • Worked with clinical leads and FLWs to map the KMC care pathway, then configured it into CommCare modules with role-based logic and conditional validations.
  • Designed Excel-based dashboards for African partner implementations, tracking risk flags, follow-up adherence, and intervention timelines.
  • Iterated workflow design with partner feedback; the Dimagi-published program has since recorded 675+ newborn enrollments and 1,400+ home visits.
Mental health CommCare on Connect Active

group Problem Management Plus (gPM+)

A digital case and operational-data system behind a group-based mental health intervention for under-served populations.

Therapy groups300+
FLWs supported~150
Supervisors~30
Program reach2,500+ clients

What it does

Frontline workers run group therapy sessions in three regions, with attendance, engagement, and outcome data captured digitally and rolled up into operational dashboards. The system lets program leadership balance workforce capacity against active caseload in something close to real time.

My role

  • Configured the case-management workflow across 300+ therapy groups, supporting ~150 frontline workers and ~30 supervisors.
  • Turned engagement and attendance data into workforce-capacity forecasts that leadership used for planning.
  • Set up quality-assurance routines that caught reporting inconsistencies before they reached leadership reports.
Child health Commodity distribution Delivered

Child Health Campaign

A digital tracking and reporting system for healthcare commodity distribution across three priority states.

Geography3 priority states
Partners5 partner organizations
FocusCommodity coverage and equity
MethodRegional uptake analysis

What it does

Tracks distribution of essential child-health commodities across three priority states, with regional uptake reporting that partner organizations use to course-correct outreach in under-served pockets.

My role

  • Improved digital tracking and performance monitoring for healthcare commodity distribution across three priority states.
  • Identified regional uptake disparities through Excel analysis, enabling targeted outreach adjustments across 5 partner organizations.

Behind the work

When the work breaks.

Moments from deployments where something went wrong, or almost did. Names of partners, platforms, and vendors are abstracted. The actions and judgment calls are mine.

Weekend ownership Root-cause tracing Resolved

A weekend bug that almost ate a pilot

A logic flaw was closing every case for a user the moment one was discontinued.

WherePilot deployment, Uganda
WhenSunday, mid-rollout
StakesActive field data collection
OutcomePilot stayed on track, no data lost

What happened

Field workers reported on a Sunday that cases were vanishing from their screens. The rest of the team was offline, and pausing for a day would have rippled into the rollout schedule.

My move

  • Traced the issue to a logic flaw: discontinuing one case was closing every case the user owned.
  • Documented the bug, asked the partner to pause data collection for the day, and looped the team in for Monday.
  • Patched the logic, reopened affected cases, and audited the same pattern across other apps in the portfolio. Caught the same flaw in two before it caused issues.
Data-quality judgment Owning the gap Resolved

Image resolution and the unverified records

A means of verification was failing by design before anyone noticed.

WhereChild-health campaign
When~1,000 records into rollout
StakesVerification of every entry
OutcomeProcess change, batch-validation now standard

What happened

Each child's vaccination was verified by an image of their immunization card. To keep field apps performant on low-end devices, image file size had been set low, but at that resolution the cards weren't readable. By the time I ran the data quality review, roughly a thousand records had already been collected without usable verification.

My move

  • Flagged the gap, paused collection, and ran a structured root-cause review with the partner.
  • Raised image resolution, tested across device tiers to confirm performance held, updated sync guidance for field users.
  • Owned the ~1,000-record gap in the post-mortem rather than papering over it. Now design against the problem by validating the data pipeline end-to-end against a small batch before scaling collection.
Pragmatic workaround Vendor management Resolved

A missing gateway

The SMS gateway didn't support the country code we needed, days before training.

WhereMulti-country rollout
WhenDays before pilot training
StakesField users couldn't create accounts
OutcomePilot started on schedule, vendor integration followed

What happened

The SMS gateway didn't support the country code we needed, so field users couldn't create accounts. The tech team was working through it from a different angle, and pilot training was already on the calendar.

My move

  • Joined the gateway as a member, monitored OTPs as they were triggered, and provided each one to field users myself so accounts never passed through me.
  • Kept the pilot on schedule with the manual workaround until a durable solution was in place.
  • Drafted the technical design for a long-term fix and coordinated onboarding of an external SMS vendor for the rest of the platform.

Process improvement

An internal project worth flagging.

Internal tooling User-centered design Adopted

Migrating the annual SWAT skill survey

From Google Forms into a CommCare app with interactive logic and dashboards on top.

ContextDimagi SWAT knowledge team
ScopeAnnual team-wide skill survey
ResultStandard tool for subsequent years
BonusStrongest capacity-building plan that cycle

What happened

The annual SWAT skill survey lived in Google Forms, which made data collection clunky and analysis even clunkier. As part of the SWAT knowledge team, I proposed moving it to CommCare, a platform we already used.

My move

  • Designed the survey inside CommCare with interactive logic and a more engaging respondent experience.
  • Built dashboards on top of the collected data so stakeholders could see skill trends across teams and levels.
  • The CommCare-based survey became the standard tool for subsequent years and fed a stronger capacity-building plan that cycle.

Independent build · 2026

AI-Powered Job
Application Tailoring System.

View source on GitHub

Job hunting in the age of AI is brutal. Hours to find roles that actually fit, then more hours editing resumes one-by-one. If résumé shortlisting on the employer side is already automated, why isn't the job-hunting side?

I'm an engineer at heart. Solving problems is my first instinct. And I'm not a developer by trade, so building this also meant learning a new skill: understanding what AI can and can't do, and how to build with it responsibly.

What it does

  1. Discovers job postings
  2. Scores match against base resume
  3. Tailors resume + cover letter for jobs scoring ≥75
  4. Renders 2-page A4 PDFs
  5. Sends a daily digest email of only high-fit matches

Stack

  • BackendPython · Flask · ~10 REST endpoints
  • Orchestrationn8n workflows in Docker
  • Data ingestionPlaywright (browser automation) · BeautifulSoup (HTML parse)
  • AIGemini 3.1 Flash Lite for scoring and tailoring
  • DocumentsWeasyPrint + Jinja2 (PDF) · pdfplumber (resume read) · openpyxl (Excel I/O)

Two ingestion modes

Automated ingestion from structured job aggregators, plus a manual URL batch mode where I curate URLs in an 8-column Excel and the same scoring/tailoring pipeline runs against each. Idempotent across re-runs via per-row status tracking, with a fetch-failed → manual-JD fallback path so the system stays useful on the days the ingestion layer breaks.

Quality controls

A 3-step structured prompt (key-term extraction → 0–100 scoring → tailored JSON output), anti-hallucination rules (banned strong-claim patterns, mandatory source-artifact naming in summaries, 4-item self-validation checklist), defensive placeholder scrubbing on Gemini output, and CSV-level deduplication via normalized URL matching.

Outputs

2-page A4 resume + cover letter PDFs per qualified job · daily digest email · complete audit log in jobs_log.csv (score, reasoning, source URL, PDF path).

How it fits together

Job aggregators
→
Playwright
+ BeautifulSoup
→
Excel batch
manual URLs
↓
n8n on Docker · Flask API
↓
Gemini 3.1 Flash Lite, 3-step prompt
↓
WeasyPrint + Jinja2
2-page PDFs
→
Gmail API
daily digest
→
jobs_log.csv
audit trail

Screenshots

A look at the bot in motion: the orchestration canvas, the daily digest, and a tailored resume sample.

n8n workflow canvas showing the bot pipeline
n8n workflow canvas. Schedule → discover → score → tailor → digest.
Daily digest email showing scored job matches
Daily digest email. Scored matches delivered each morning.
Tailored resume PDF generated by the bot
Tailored resume PDF. Generated per job from a master resume.

Lessons learned

  1. Anti-hallucination is a design problem, not a prompt-tuning one. Models will quietly relabel candidate experience to match job-description language. Banned-pattern checks plus a self-validation pass caught what prompting alone could not.
  2. Narrow LLM responsibility produces better output. Splitting the job into discrete steps (extract key terms → score → tailor) returned more reliable JSON than asking one prompt for everything.
  3. AI-assisted development is real, but not magical. I leaned on Claude to write code I couldn't write from scratch, and still had to learn enough Docker, n8n, and Playwright to debug what it produced.

Hosting & privacy

Runs locally on-device. Not deployed publicly, by design, since it handles personal Gmail credentials and unfiltered job data. The architecture is portable and could be containerized for cloud deployment if a multi-user need arose.

Education

Degree and continuing coursework.

Formal degree, then a working timeline of certifications, most recent first.

2017 to 2021

Bachelor of Technology, Electronics & Communication Engineering

Poornima University, Jaipur, India

Certifications & coursework

  1. May 2026

    Career Essentials in Project Management

    Microsoft · LinkedIn Learning

  2. Apr 2026

    Project Management Foundations

    Project Management Institute

  3. Apr 2026

    Atlassian Agile Project Management Professional Certificate

    Atlassian

  4. Apr 2026

    Agile Foundations

    LinkedIn Learning

  5. Mar 2026

    Data Analysis using Microsoft Excel

    Coursera

  6. Nov 2022

    Business Analysis & Process Management

    Coursera

  7. Nov 2022

    Get Started with Jira

    Coursera

  8. Aug 2022

    Foundations of Project Management

    Coursera

  9. Aug 2020

    Programming for Everybody (Python)

    Coursera

  10. Jun 2020

    Introduction to Electronics

    Coursera

  11. May 2020

    Introduction to the Internet of Things & Embedded Systems

    Coursera

  12. Apr 2020

    Industry 4.0

    Tata Steel

  13. Apr 2020

    Fundamentals of Digital Marketing

    Google

  14. May 2020

    English Proficiency (C1)

    British Council