Manish Pithwa
AI Engineer
19+ years exp
Management: 10+ years (50+ people)
JavaScript
Python
TypeScript
ABAP
Defensive Hardening: 5/10
Active 6 days ago
Invite to interview
Download CVCV
Overview
Technical skills
Timeline
Roles
Overview
Application Security-oriented web developer (senior-level engineering signal) focused on building production web apps with integrated LLM services and defensive controls. The strongest proven skill is building resilient AI-backed form-generation and repair logic with concrete artifacts in app/Services/AiFormService.php and comprehensive feature tests in tests/Feature/FormBuilderTest.php. The public code lacks formal offensive research, cryptography/key lifecycle engineering, and mature detection/IR artifacts or formal threat models.
Phone
Technical skills
Languages
4
JavaScript
Python
TypeScript
ABAP
DevOps
12
AWS
Amazon S3
Amazon EC2
DigitalOcean
Nginx
Ubuntu
Bitbucket
Docker
GitHub
Azure
Kubernetes
GCP
Databases
7
ElasticSearch
PostgreSQL
MariaDB
Redis
SAP HANA
MySQL
Oracle
Frontend
4
Angular
Vue.js
React.js
Next.js
AI/ML
17
LoRA
OpenAI
Embeddings
RAG
Agentic Workflows
LLM
Claude
CrewAI
Gemini
LangChain
Llama
LlamaIndex
PyTorch
Scikit-learn
Sentence-Transformers
TensorFlow
Transformers
Other
42
OpenSSL
OpenSSH
MVC
Playwright
SAP BTP
Jest
Asyncio
PEFT
Semantic Search
Grafana
Hugging Face
MLFlow
AppSec
Cybersecurity
Terraform
Knowledge Distillation
Microsoft Entra ID
Docker Compose
Ansible
Prometheus
Model Context Protocol
CI/CD
Jenkins
AI Agents
GGUF
Machine Learning
QA
WebRTC
WebSockets
Debian
Prompt Engineering
Fine-tuning
Function Calling
NLP
Structured Outputs
Text-to-Speech
Voice Agents
Quantization
Diffusion Models
Model Distillation
Realtime API
Sentiment Analysis
Timeline
Senior AI and Backend Developer
•
Senior
MICRONISUS SYSTEMS
•
Full-Time
Designed Python-based LLM application workflows that combined RAG, vector search, and agentic orchestration for knowledge retrieval and multi-step execution. Evaluated model adaptation and optimization approaches such as LoRA, quantization, and knowledge distillation while applying security-conscious API design. Built cloud deployment patterns and implemented observability using Prometheus, Grafana, and ELK-style log analytics, integrating Azure AI services into backend workflows.
Python
Agentic Workflows
RAG
LLM
OpenAI
Azure
Embeddings
Semantic Search
LoRA
Quantization
Knowledge Distillation
Prometheus
Grafana
ElasticSearch
Docker
Kubernetes
CI/CD
Jenkins
Terraform
Ansible
Microsoft Entra ID
AWS
GCP
PostgreSQL
MySQL
Oracle
Redis
RabbitMQ
MLFlow
Agile
Senior PHP Developer
•
Senior
FIRST DUE SOLUTIONS INDIA PVT. LTD.
•
Full-Time
Developed and maintained Yii/Vue.js applications and backend/API integrations. Worked across frontend and backend layers, using containerized development and database/messaging components to support application delivery. Supported cloud-hosted environments with Ubuntu and Nginx and used GitHub-based workflows for source control and release-related activities.
PHP
Yii
Vue.js
Docker
AWS
Ubuntu
Nginx
PostgreSQL
RabbitMQ
Redis
GitHub
Head Of Technology
•
Head+
PINKNBLU INFOMEDIA PVT. LTD.
•
Full-Time
Led a cross-functional technology team, coordinating delivery activities across engineering, QA, and product execution. Managed development and product initiatives using Laravel and Angular, including work that involved Elasticsearch-based search and delivery planning. Oversaw AWS deployments across EC2 and S3 and aligned requirements, estimation, and release governance across parallel streams.
Laravel
Angular
ElasticSearch
Amazon EC2
Amazon S3
AWS
Lead Fullstack
•
Lead
INTENSE TECHNOLOGIES LIMITED
•
Full-Time
Provided technical leadership for a team delivering PHP and Laravel-based solutions with Angular frontends. Owned client requirements clarification, estimation, and end-to-end delivery coordination for multiple projects. Contributed to the delivery process using a stack centered on Laravel, WordPress, and Angular.
PHP
WordPress
Laravel
Angular
Project Manager / Self Entrepreneur
•
Lead
SELF ENTREPRENEUR
•
Full-Time
Built websites, web services, and REST/API integrations using Laravel/Lumen with Angular frontends. Deployed applications on DigitalOcean using Ubuntu and Nginx, and used MariaDB for data storage. Owned client solution delivery from requirements through implementation, deployment, and ongoing support handover.
Laravel
Angular
Rest API
DigitalOcean
Ubuntu
Nginx
MariaDB
Bitbucket
Sr. Team Lead
•
Lead
LAXYO SOLUTION SOFT PVT. LTD.
•
Full-Time
Led development teams responsible for maintaining and delivering web applications with PHP-based frameworks and CMS platforms. Coordinated delivery from client requirements through implementation, deployment, and post-release follow-ups with an emphasis on maintainability. Supported execution across Laravel, Yii-based components, and WordPress implementations.
Laravel
Yii
WordPress
Team Lead
•
Lead
CYBER INFRASTRUCTURE PVT. LTD.
•
Full-Time
Led PHP application development efforts involving JavaScript and MVC-style architectures. Contributed to database design, code reviews, and modular object-oriented implementation while coordinating with clients on delivery needs. Supported WordPress development and quality practices across the team.
PHP
JavaScript
MVC
WordPress
Team Lead
•
Lead
INTERCOM ONLINE PVT. LTD.
•
Full-Time
Worked on PHP and JavaScript development in a LAMP environment, including MVC-style module work. Supported database design and code reviews while coordinating team execution and client interactions. Contributed to Joomla-based implementations alongside PHP MVC development.
PHP
JavaScript
MVC
Rajiv Gandhi Proudyogiki Vishwavidyalaya (Rajiv Gandhi Technical University)
Associate's Degree •
Information Technology
Senior Security Engineer
Confidence: Medium AppSec Engineer
Application Security-oriented web developer (senior-level engineering signal) focused on building production web apps with integrated LLM services and defensive controls. The strongest proven skill is building resilient AI-backed form-generation and repair logic with concrete artifacts in app/Services/AiFormService.php and comprehensive feature tests in tests/Feature/FormBuilderTest.php. The public code lacks formal offensive research, cryptography/key lifecycle engineering, and mature detection/IR artifacts or formal threat models.
Vulnerability Analysis
3/10
Finding security weaknesses
Basic vulnerability analysis and defensive tests are present (SQLi/XSS checks and schema-repair logic) but there is no deep root-cause exploitability research, affected-version analysis, or public CVE-level work.
Evidence
new-aggregator-api/tests/Feature/SecurityTest.php:test_sql_injection_protection
form-builder-ai/app/Services/AiFormService.php:repairSchema and repairTruncatedJson functions (defensive handling of malformed LLM JSON)
Cryptography & Secure Protocols
1/10
Using encryption correctly
Minimal cryptography/protocol engineering; API tokens are used but there is no evidence of key lifecycle management, TLS/mTLS configuration, or protocol misuse tests.
Evidence
form-builder-ai/app/Services/AiFormService.php: uses Http::withToken() to call LLM providers
docker-compose.yml: services expose internal ports without TLS termination config
Offensive Security / Pentesting
1/10
Testing by attacking
No dedicated offensive security artifacts or PoCs; only defensive tests that assert protections exist (e.g. SQLi/XSS protections) are present.
Evidence
new-aggregator-api/tests/Feature/SecurityTest.php:test_sql_injection_protection
Defensive Hardening
5/10
Making systems harder to break
Clear defensive hardening work: security middleware, optimized rate limiting, server-side validation, caching/invalidation and tests; shows considered defensive design though some deployment defaults are insecure.
Evidence
new-aggregator-api/app/Http/Middleware/SecurityHeaders.php
new-aggregator-api/app/Http/Middleware/OptimizedRateLimiting.php
new-aggregator-api/tests/Feature/SecurityTest.php (headers, CORS, rate limiting checks)
Detection & Incident Response
1/10
Spotting and handling attacks
Little to no detection/IR engineering shown; there are tests asserting protections but no SIEM, Sigma/YARA rules, ATT&CK mapping, or runbooks.
Evidence
new-aggregator-api/tests/Feature/SecurityTest.php:test_rate_limiting_on_sensitive_endpoints
Compliance & Threat Modeling
2/10
Planning against threats
Some application-level multi-tenant awareness and migrations exist, indicating elementary design-for-separation; there is no formal threat modeling, STRIDE/attack-trees or compliance documentation.
Evidence
form-builder-ai/app/Traits/BelongsToTenant.php
form-builder-ai/database/migrations/2026_08_09_161206_add_tenant_id_to_tables.php
Expertise
Application Security• Middle
Industries
Software• Middle
Technologies
Cybersecurity
AppSec
Microsoft Entra ID• since 2025
Python• mentioned only
Recommendations
- Develop secure web applications that integrate LLMs and streaming APIs, e.g. build LLM streaming endpoints, robust fallback paths and schema repair logic.
- Implement hardened container images and CI secrets handling - remove chmod 777 usage and adopt least-privilege file permissions and proper secrets management (Vault/ENV policing).
- Expand security engineering work into detection and IR: add Sigma/Suricata rules, ATT&CK mapping, and reproducible PoCs for critical issues with regression tests.
- Harden deployment configs: remove hardcoded default credentials in docker-compose, add TLS termination (or mTLS for internal services), and add automated IaC security scans.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Middle Backend Developer
Confidence: Medium API Engineer
Senior API-focused backend engineer specializing in building AI-integrated web services and developer-facing workflows. The strongest proven skill is designing resilient LLM integration and form-generation streaming—evidenced by app/Services/AiFormService.php (SSE streaming tiers, fallback repair, caching) and comprehensive integration tests in form-builder-ai/tests/Feature/FormBuilderTest.php. Public code does not show deep distributed-systems work, raw SQL performance tuning, or explicit cloud infra/operations automation artifacts.
API Design
5/10
How well APIs are designed
API design shows deliberate choices for streaming (SSE), multi-tier fallbacks, and an interface-driven AI service, but lacks explicit versioning, idempotency key patterns, and a formal error-code contract across services.
Evidence
form-builder-ai/app/Services/AiFormService.php: streamFormGeneration SSE tiers and direct LLM streaming fallback
form-builder-ai/app/Contracts/AiServiceInterface.php: explicit API contract used by Laravel services and controllers
form-builder-ai/ai-service/main.py: Python FastAPI endpoints referenced as Tier 1 microservice (fastapi / stream-generate-form)
Data Layer & Database
4/10
Working with databases
Data layer is well-structured with migrations, seeders, models and integration tests, but there is limited evidence of advanced transaction handling, explicit isolation level control or hand-tuned SQL.
Evidence
form-builder-ai/database/migrations/2026_08_05_000001_create_forms_table.php: migration history and schema evolution
form-builder-ai/app/Models/Form.php: Eloquent model with lifecycle boot logic
form-builder-ai/tests/Feature/FormBuilderTest.php: RefreshDatabase-backed integration tests exercising persistence
Scalability & Performance
4/10
Handling load and speed
Shows practical scalability work: SSE streaming, Redis caching, queued background jobs and ability to toggle queued imports; lacks load-testing artifacts or measured tuning evidence.
Evidence
form-builder-ai/app/Services/AiFormService.php: Cache::remember usage and chunked SSE fallback streaming
form-builder-ai/app/Jobs/GenerateFormSchemaJob.php and ParseDocumentJob.php: queued background job implementations
form-builder-ai/resources/views/livewire/document-importer.blade.php: UI option and polling for queued background import mode
System Architecture
4/10
Overall system structure
Reasonable system decomposition and separation of concerns (Laravel UI, AI microservice, job queue, storage abstraction), but no evidence of multi-service contracts, service discovery, or distributed-system tradeoffs beyond basic decoupling.
Evidence
form-builder-ai/app/Providers/AppServiceProvider.php: service bindings for AiServiceInterface and storage/document parser abstractions
form-builder-ai/ai-service/pyproject.toml and main.py: presence of a separate FastAPI microservice
form-builder-ai/app/Livewire/FormBuilder.php: client-side/Livewire component interacting with service layer and jobs
Security & Auth
3/10
Protecting data and access
Basic security hygiene and input validation are present (server-side schema validation, honeypot test, file upload handling), but there is limited public evidence of token lifecycle management, secrets rotation, or dependency auditing integrated into runtime code.
Evidence
form-builder-ai/app/Livewire/PublicFormFill.php (routes referenced): server-side dynamic validation and honeypot protections exercised in tests
form-builder-ai/tests/Feature/FormBuilderTest.php: tests for honeypot trapping and file upload/download using Storage::fake
form-builder-ai/app/Services/AiFormService.php: checks for placeholder API keys and uses Http::timeout with try/catch for external calls
Reliability & Observability
4/10
Stability and monitoring
Observability and reliability patterns are implemented at the application level (generation logs, try/catch fallbacks, timeouts), but advanced resilience patterns like configured backoff retries, circuit breakers, and distributed tracing configuration are limited in the human-authored artifacts.
Evidence
form-builder-ai/app/Services/AiFormService.php: AiGenerationLog entries, try/catch fallbacks and Http::timeout usage
form-builder-ai/tests/Feature/FormBuilderTest.php: assertions verifying ai_generation_logs are created and status updated
form-builder-ai/app/Jobs/*: background job patterns and use of queueing indicate asynchronous processing for reliability
Expertise
PHP• Middle
Backend AI & LLM• Middle
Python• Middle
Recommendations
- Lead development of API-first SaaS features that integrate LLMs and streaming responses (use the AiFormService streaming + FastAPI pattern).
- Build developer-facing microservices or SDKs that expose AI-backed schema generation and robust fallback/retry semantics.
- Implement production-grade observability and resilience hardening: add structured tracing, configured retry/backoff policies, and load testing for SSE endpoints.
- Own form/tenant data integrity work and schema evolution: design migrations, transactional boundaries and explicit isolation behavior for multi-tenant tenants.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Junior QA Engineer
Confidence: Medium Automation Engineer
Automation-focused engineer at a Middle level specializing in actionable end-to-end and unit test engineering. The strongest proven skill is building resilient HTTP and integration client logic with adaptive network handling demonstrated in shared-household-chores/frontend/src/services/httpClient.ts. There is limited public evidence of contract testing, large-scale performance/load testing, CI test-sharding, or documented RCA-driven bug triage.
Test Automation Frameworks
5/10
Building automated tests
Test automation frameworks in use - Playwright demo scripting and Vitest unit test config are present with reasonable selector handling, retries and environment awareness, though some demo scripts rely on sleeps rather than fully guarded waits.
Evidence
shared-household-chores/e2e/demo/record.js
shared-household-chores/frontend/src/services/httpClient.ts
todo-app/vitest.config.ts
Test Coverage & Strategy
4/10
What and how to test
Test coverage & strategy shows meaningful unit and component tests with edge-case assertions and SSR/localStorage safety checks, but there is no visible risk-based test tagging, property-based tests, or mutation/coverage honesty artifacts in human-authored files.
Evidence
todo-app/src/types/index.test.ts
todo-app/src/lib/db/repository.test.ts
API & Integration Testing
3/10
Testing how parts work together
API and integration testing competency is evident in a resilient HttpClient implementation and fallback patterns, but there is no human-authored contract testing (Pact/schemathesis) or integration testcontainers evidence.
Evidence
shared-household-chores/frontend/src/services/httpClient.ts
todo-app/src/lib/ai/tools.ts
Performance & Load Testing
1/10
Testing speed under load
Performance and load testing is not evidenced; there are timed demo flows and scripted waits but no load scenarios, thresholds, or automated percentile analysis.
Evidence
shared-household-chores/e2e/demo/record.js
Bug Reporting & Analysis
1/10
Finding and describing bugs
Bug reporting and root-cause analysis artifacts are minimal; code includes error handlers and defensive checks but no reproducible bug reports, issue-to-fix linking, or flaky-test RCA documentation in human-authored files.
Evidence
courier-service/courier_common.py
courier-service/courier_delivery_time.py
CI Test Integration
2/10
Running tests automatically
CI/test integration artifacts are minimal in human-authored files; there is a Vitest config and Playwright demo script oriented for CI demos but no matrix, per-test artifacts retention policies, selective runs on diff, or quarantine mechanics visible.
Evidence
todo-app/vitest.config.ts
shared-household-chores/e2e/demo/record.js
Expertise
SDET & Test Engineering• Junior
Unit & Component Testing• Junior
Industries
Lifestyle• Junior
Technologies
QA
TypeScript• Junior
Playwright
Docker Compose
Jest
Vitest• mentioned only
Recommendations
- Expand contract and integration testing with Pact or schemathesis to validate OpenAPI behavior end-to-end and add schema-driven negative-path tests.
- Add CI test matrix, per-test artifact retention, and per-test quarantine or targeted retries instead of global retries to make E2E runs reliable in CI.
- Introduce mutation testing or stronger coverage gates and property-based tests for critical invariants (eg. auth flows, point accounting) to increase test honesty.
- Replace sleep-based waits in demo scripts with condition-based waits and explicit flaky-test handling to reduce nondeterminism in E2E suites.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories: