Experience

• 

  Lahore electric supply company limited

  Jun 2019 - Aug 2019

  Internship trainee
• 

  Ptcl.official

  Oct 2020 - Dec 2020

  Internee

  Three months in telecom power and switching. Visited sites, fixed the power issues, routing the IPs, and handling customers through CRM. Experienced a lot with interpersonal skills.

• 

  Cybernet

  Mar 2021 - Aug 2022

  Data analyst

  Field work -> FTTH deployment (from Datacenters to Splitter in PON architecture)---Managing GIS data, designing dashboards and charts (Excel, Tableau, QGIS)Worked on PostGIS, GRASS and graph data structures to systematize GIS dataAnalysis and operation optimization of GIS data, for optimal planning (Python, R, MATLAB)Detecting and fixing the erroneous database entries of heirachical GPON data -> a single base node NULL/wrong entry could cause a loss of information of all its leaves. Used KNN to solve that. ----Coordinated patrolling teamDeveloped a local DB and basic Access frontend design for automation in operations of data entry Show less

• 

  Creditper

  Aug 2022 - Apr 2023

  Machine learning engineer

  Backend development, managing APIs and Fixing Bugs[Projects][CV] Automatic document cropper with information validation using OCR.[NLP] Use BERT on SMS data for classification → generate user income flows. Develop an API to request this model to generate income flow of a particular userID

• 

  National university of computer and emerging sciences

  Feb 2023 - Jul 2023

  Research assistant

  Worked on the design of Magneto Inductive (MI) Communication as a part of FRSG project assigned to MARS Lab.Engineered the communication model and develop simulations to transmit magnetically encoded information through ASK. Automates the processes through software and instrument control. Design STM32 based sensor nodes, with a single node could sent analog information and decode it.

• 

  Jazzcash

  Aug 2023 - now

  Product specialist - data analyst

  Develop ETL jobs using spark, teradata sql, shell scripts, talend and python for report automation. Independently developed a reporting (dashboarding) and ELT ecosystem with bare python to flexibly meet the requirements. And develop it raw using the best DE practices. Used airbyte to do CDC on EL layer and DBT for transformation. Setup psql, and oracle databases, connect them to lake to perform better analytics. The pipelines develop can manage GBs/hr, enough for our needs. Even for that I had to use python threading.Develop the daily email reports to flexibly analyze the lending, savings and insurance. I was also involved with projects including NBFC reporting to help bank recon. Finding problems in loan reporting calculations and managing vendors for fixation. Using the developed DE system, I have worked with various analytics for product growth including:1. Target segmentation of customers using unsupervised clustering techniques. 2. Working with lending team to minimize default risk. With the team effort over the period of year we were able to grow digital lending to 130% while keeping the default rate same. Majorly I have worked with nano-lending, savings, insurance and BNPL. I did analytics with product owners to achieve growth targets. For analytics I have mostly worked with inferential and baysian statistics to meet the needs. Some studies which I conducted over the period:1. Selecting top merchants for over the limit disbursement (used descriptive statistics) - boosting growth2. Strategize to increase flow of competitors' (can't name) customer on JC loans. 78% customers coming for competitor's repayment shifted. 3. Risk profiling for RRBT (recovering loans)Currently for feature intensive datasets I am also working on developing DL systems. Currently using DataIKU, but also developing system on torch for flexibility. Here I am trying to use MLOps practices, and developing containerized applications using Garaj on-premise machines. Show less