my_codes-and-project/project_Done at master · yashugupta786/my_codes-and-project · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
EXL
Neural translation
Tesseract engine for converting  pdf to editable word
Image processing engine
classification email,other embeded analytics
corona analysis from different sources and making wordcloud for that using custom training on fast ai and word2vec
#------------------------------------------------
Eclerx
NMT(Seq2Seq with attention)
Tia ticket classification (laptopuser/ISG/HR/Compliance/RDP/windows/password/reset/tools)
S4x similarity using elmo and word vectors   client Dentions
NPS predictor   COMCAST
R2R Predictor
NPS Drivers  COMCAST
BAXA Classification (Demonstarive,ornot demonstation),binary text and multiclass prediction Dentions
Chatvolume forcast (in which we have the original dates and actuals calls for different vertical like billing,reconsilation)
Nps prediction and predicting the classes
churn prediction
admin bot
#---------------monster projects ----------------
1 chatbot
NER
Tesseract Resume parser
Classification using RNN for classifyng extracted content from resume
recommendation engine
semantic search
intent classification
GEnpact
Intelligent connectiion
neural Platform

Future projects
S4x similarity auto taging
We have legal client in which a 15 min call happens and based on that call bpo agents answer several questions by entering a string.The questions are same and they need to answer that
questions based on the call.They need to input various strings
That string needs to be validate by most of times agent enters same string which reduces the productivity so we implemented a engine fo fiding the similarity of text they enter
We match the string using elmo embeding and finding the similarity and
Find the score if score is less than 60 percent we take string is unique
And agent can enter that string


After that we devloped a engine which classify the string enter by BPO agents  into different category that is demonstrated or no demonstrate or partial demonstrate. This action is
done by the audit team .... with same data  s4x there is baxa tag
Using bert language model
 After the user submission
There is aaudit team who audit the answers entered by user
Into demonstrate,not demonstrate,partial demonstrate
To find the quality of agents and the  call feedback by user