DP divyajeetpanda@outlook.com
Resume

Hello ! I'm

Divyajeet

Software Engineer · Vimaan Robotics · 3+ years

A software engineer for

backend, data and graph systems

systems

About me

I build backend and data systems. At Vimaan Robotics I write the Cypher and SQL that turn computer-vision scans into live warehouse inventory. When those numbers don't match the source records, my checks are what flag it. Before that I spent over two years at Cognizant building Python and Node.js services.

Away from the keyboard I shoot photos and video, read fiction and fantasy, and travel whenever I can.

What
I do

Backend & data

Services, queries, pipelines

Java/Spring Boot and Python services on PostgreSQL and Neo4j, and the Cypher and SQL behind live client dashboards.

Skills, and where I've used them

JavaSpring Boot
Vimaan Robotics
Node.js
Cognizant
KafkaRabbitMQ
Cognizant
Kubernetes
Cognizant
CI/CDDeploymentsLinux
Vimaan Robotics
Pipeline configurationFrontend configuration
Vimaan Robotics
Grafana
Cognizant

Graph & AI

LLMs, agents, fine-tuning

Text-to-Cypher with local models, so answers come from the data instead of the model's memory. Agents on Microsoft Foundry and Copilot Studio. Structured output from small local models, with schema checks and a LoRA training pipeline.

Skills, and where I've used them

OllamaStreamlit
Olist Query Studio
PyTorchLoRA / PEFT
Bug2JSON

How it connects

Pick a skill to see where I've used it. Pick a job, project or certificate to jump to it.

The same links are listed under each skill in the section above.

Where I've
worked

Software Engineer

Vimaan Robotics · Bengaluru

NOWAug 2025 – Present

  • Cut new-warehouse onboarding from 2–3 days to under 1 day with a Python configuration builder and UI.
  • Configure facility-specific Cypher and SQL that turn computer-vision scans into live inventory on client dashboards.
  • Build reconciliation logic that flags mismatches, missing entries, and exceptions against source records.
  • Debug Java/Spring Boot services, graph pipelines, and dashboards; support releases with log analysis and regression testing.
  • Run CI/CD and Docker-based deployments on Linux, and maintain pipeline and frontend configuration.

Senior Systems/Software Engineer

Cognizant · Bengaluru

2022Sep 2022 – Jan 2025

  • Cut PostgreSQL query execution time by 25% and improved overall query performance by 19%, earning the Cognizant Excellence Award.
  • Shipped Python/FastAPI and Node.js microservices on Docker and Kubernetes across AWS and Azure.
  • Built REST and GraphQL APIs with OAuth 2.0/JWT, rate limiting, and caching, plus Kafka/RabbitMQ event-driven workflows.
  • Set up observability with Prometheus, Grafana, and ELK; mentored junior engineers through code reviews.

Personal projects

01

Olist Query Studio

LLM · Graph analytics

Turns plain-English analytics questions about the Olist e-commerce dataset into Cypher and runs them against Neo4j.

What I built

  • Query generation from the graph schema with a local Ollama model
  • A validator, an EXPLAIN check and a read-only account on every query

Matches 82 of 112 verified benchmark questions (73%)

How it flows

  1. Plain-English question
  2. Local Ollama modelwrites Cypher from the graph schema
  3. Validator + EXPLAINchecked before it runs
  4. Neo4jread-only account
  5. Answer

Try a question

How many orders were delivered?

Recorded run · Cypher written by the local model

MATCH (o:Order {status: 'delivered'}) RETURN count(o) AS deliveredOrders
deliveredOrders
96,478

Which five translated product categories generated the most delivered-order item GMV, excluding freight?

Benchmark · checked query and answer

MATCH (:Order {status: 'delivered'})-[:HAS_ITEM]->(i:OrderItem)
      -[:OF_PRODUCT]->(:Product)-[:IN_CATEGORY]->(c:Category)
RETURN c.nameEnglish AS category,
       round(sum(i.price) * 100) / 100 AS deliveredItemGmvBrl
ORDER BY deliveredItemGmvBrl DESC, category
LIMIT 5
categorydeliveredItemGmvBrl
health_beauty1,233,131.72
watches_gifts1,166,176.98
bed_bath_table1,023,434.76
sports_leisure954,852.55
computers_accessories888,724.61

What is the average review score for delivered orders that arrived after their estimated delivery date?

Benchmark · checked query and answer

MATCH (o:Order {status: 'delivered'})-[:HAS_REVIEW]->(r:Review)
WHERE o.deliveredAt > o.estimatedDeliveryAt
RETURN avg(r.score) AS averageReviewScore
averageReviewScore
2.57

Which five customer states have the most unique customers?

Benchmark · checked query and answer

MATCH (c:Customer)
RETURN c.state AS state, count(DISTINCT c.uniqueCustomerId) AS customers
ORDER BY customers DESC, state
LIMIT 5
statecustomers
SP40,302
RJ12,384
MG11,259
RS5,277
PR4,882

Delete all canceled orders.

Recorded run · the model declined

Declined as unsafe. “This request asks to delete data, which is not allowed by this tool. It can only read data.” No query was written or run.

Nothing runs in your browser: these are saved from the real tool and its benchmark.

The technical detail

Technologies

  • Python
  • Neo4j
  • Cypher
  • Ollama
  • FastAPI
  • Streamlit

Tradeoffs

A local 7B model keeps the data on the machine, but an answer takes several seconds and matches 73% of the benchmark. One correction attempt keeps latency bounded; the validator is a second check, and the read-only database account is what actually stops writes.

View code

02

Bug2JSON

LLM · Bug triage

Turns free-text bug reports into schema-validated triage records, from category and severity to a suggested next step.

What I built

  • A FastAPI service running a local Qwen 2.5 model
  • Guardrails that mark anything the report doesn't support as “unknown”
  • A React review UI that keeps the JSON in view

30 of 30 benchmark responses schema-valid

How it flows

  1. Free-text bug report
  2. FastAPI + Qwen 2.5runs locally
  3. Guardrailsunsupported claims become “unknown”
  4. Schema check
  5. React review UI
The technical detail

Technologies

  • Python
  • PyTorch
  • LoRA / PEFT
  • FastAPI
  • React
  • TypeScript

Tradeoffs

Every response is schema-valid (30 of 30), but the 1.5B base model’s labels are weak on the synthetic suite (40% category accuracy), so the guardrails prefer “unknown” and the UI asks for review. No fine-tune result is claimed until a reviewed, real-world evaluation exists.

View code

03

Sort-Downloads

Python · CLI · Automation

A command-line tool I use daily that files every loose download by category, extension and month.

What I built

  • Filing rules by category, extension and month
  • Exact-duplicate detection by size and MD5, sent to the Recycle Bin
  • A colour-coded preview of every move before it runs

Parity test: files everything exactly as the PowerShell original

How it flows

  1. Downloads folder
  2. Sort rulescategory, extension, month
  3. Duplicate checksame size and MD5
  4. Colour-coded preview
  5. Moveduplicates to the Recycle Bin
The technical detail

Technologies

  • Python
  • Rich
  • PowerShell
  • pytest

Tradeoffs

Duplicates go to the Recycle Bin, grouped in one folder so they restore in one go, and name clashes become “name (2)” instead of overwriting. Matching on size plus MD5 only catches byte-identical copies, which is what a Downloads folder needs.

View code

04

AI Resume Analyzer

Full stack · AI

Compares PDF/DOCX resumes with job descriptions and returns ATS-style feedback, keyword gaps and suggestions.

What I built

  • A Next.js frontend and an Express/TypeScript API that share Zod schemas
  • Supabase auth
  • OpenAI-assisted feedback backed by deterministic scoring

In progress · repository not public yet

How it flows

  1. Resume + job descriptionPDF or DOCX
  2. Next.js UISupabase auth
  3. Express/TypeScript APIZod schemas shared with the UI
  4. Scoringdeterministic, with OpenAI-assisted feedback
  5. ATS-style report
The technical detail

Technologies

  • Next.js
  • TypeScript
  • Express.js
  • Supabase
  • OpenAI API
  • Zod

Tradeoffs

OpenAI writes the feedback, but the score comes from deterministic scoring, with a heuristic fallback when no API key is set. Sharing Zod schemas between the Next.js app and the Express API keeps both sides on one contract.

05

Ethereal Command

System telemetry dashboard

A local system-monitoring dashboard for CPU, RAM, GPU, storage, thermal, network and process telemetry.

What I built

  • An Express backend that polls telemetry behind cached endpoints
  • A React/Vite UI with live cards, time-series charts and process tables

In progress · repository not public yet

How it flows

  1. System telemetryCPU, RAM, GPU, storage, thermal, network, processes
  2. Express poller
  3. Cached endpoints
  4. React/Vite UIlive cards, charts, process tables
The technical detail

Technologies

  • React
  • TypeScript
  • Vite
  • Tailwind CSS
  • Express.js
  • Recharts

Tradeoffs

The backend polls on its own timers (system every second, processes every two, GPU every three) and answers from an in-memory cache, so the UI never waits on a slow sensor, at the cost of data a few seconds old.

Everything else I've built, including experiments and older work, is on GitHub. See all on GitHub

Certifications

Each credential links to the issuer's page, so you can verify it yourself.

Applied skill · Lab-assessed

Enhance agents with autonomous capabilities

Microsoft · Oct 2026

ID E2571D68A27B29EE

Show credential

Applied skill · Lab-assessed

Get started developing agents in Microsoft Foundry

Microsoft · Oct 2026

ID 9DF07DC5B9B7E5C8

Show credential

Professional certification

Neo4j Certified Professional

Neo4j GraphAcademy · Sep 2026

ID 46327d32-4193-4fb0-aa4a-783195f258bb

Show credential
First page of Divyajeet Panda's resume

Resume

Two pages covering my experience, projects, education and certifications.

  • 3+ years in backend & data systems
  • Java/Spring Boot · Python · Neo4j · PostgreSQL
  • Gold Medalist, M.Sc. IT
Download PDF
View online

PDF · 2 pages · 10 KB · Updated Oct 2026
Short link: divyajeet.netlify.app/resume

Education & awards

Education

  • 2023 – 2025

    M.Sc. in Information Technology

    Ravenshaw University, Cuttack

    CGPA 8.20 · Gold Medalist

  • 2019 – 2022

    B.Sc. in Information Technology and Management

    Ravenshaw University, Cuttack

    CGPA 8.68

Honors & awards

  • Gold Medalist, M.Sc. IT

    Ravenshaw University; medal presented by the President of India at the university convocation.

    Medal recipients with the chief guests at Ravenshaw University's 13th convocation The first-position M.Sc. IT gold medal, engraved with my name, outside the convocation venue

  • Cognizant Excellence Award

    For cutting PostgreSQL query execution time by 25%.