We are currently in open beta. Please sign up for try it for free.

We are currently in open beta.

Please sign up for try it for free.

Products

Intent Router

Filter Extractor

Pricing

Contact Us


Products

Intent Router

Filter Extractor

Pricing

Contact Us


Filter Extractor

Filter Extractor

Filter Extractor

Extracts structured filters from the query for the given field

Extracts structured filters from the query for the given field

> = 2024-01-01 AND < = 2024-12-31

Filter Extractor

Published

date

“Find articles about the impact of

Large language models on the economy

published in 2024.”

> = 2024-01-01 AND < = 2024-12-31

Filter Extractor

Published

date

“Find articles about the impact of

Large language models on the economy

published in 2024.”

> = 2024-01-01 AND < = 2024-12-31

Filter Extractor

Published

date

“Find articles about the impact of

Large language models on the economy

published in 2024.”

What does the Filter Extractor do?

What does

Filter Extractor do?

What does the

Filter Extractor do?

Filter Extractor receives a natural language query along with a field and then generates structured filters.

Filter Extractor receives a natural language query along with a field and then generates structured filters.

import sineps

client = sineps.Client(SINEPS_API_KEY)

query = "Find articles about the impact of Large language models on jobs and the economy, published in 2024"

field = {
    "name": "published_date",
    "type": "date",
    "description": "The date the article was published online.",
}

response = client.exec_filter_extractor(query=query, field=field)

print(response.result)

# {
#   "type": "ConjunctedFilter",
#   "conjunction": "AND",
#   "filters": [
#     {
#       "Filter": {
#         "type": "Filter",
#         "operator": ">=",
#         "value": "2024-01-01"
#       }
#     },
#     {
#       "type": "Filter",
#       "operator": "<=",
#       "value": "2024-12-31"
#     }
#   ]
# }

Values

Values

Low-cost

Low-cost

164x cheaper than GPT-4o,

5x cheaper than GPT-4o mini

164x cheaper than GPT-4o,

5x cheaper than GPT-4o mini

Average cost / 1K calls in filter extraction problems

GPT - 4o

mini

GPT - 4o

GPT - 3.5

turbo

Claude 3.5

Sonnet

0.5$

1.6$

Sineps

11.6$

16.4$

0.1$

16 $

12 $

8 $

4 $

0 $

Average cost / 1K calls in filter extraction problems

GPT - 4o

mini

GPT - 4o

GPT - 3.5

turbo

Claude 3.5

Sonnet

0.5$

1.6$

Sineps

11.6$

16.4$

0.1$

16 $

12 $

8 $

4 $

0 $

Consistently Fast

Consistently Fast

Average latency 5x lower,

Maximum latency 12x lower than GPT-4o mini

Average latency 5x lower,

Maximum latency 12x lower than GPT-4o mini

  • Average latency in filter extraction problems

    GPT - 4o

    mini

    GPT - 4o

    GPT - 3.5

    turbo

    Claude 3.5

    Sonnet

    Sineps

    655ms

    1720ms

    810ms

    830ms

    160ms

    1800ms

    1350ms

    900ms

    450ms

    0ms

  • Maximum latency (at 300 calls) in filter extraction problems

    GPT - 4o

    mini

    GPT - 4o

    GPT - 3.5

    turbo

    Claude 3.5

    Sonnet

    Sineps

    1350ms

    3547ms

    9100ms

    4900ms

    382ms

    9000ms

    6750ms

    4500ms

    2250ms

    0ms

  • Average latency in filter extraction problems

    GPT - 4o

    mini

    GPT - 4o

    GPT - 3.5

    turbo

    Claude 3.5

    Sonnet

    Sineps

    655ms

    1720ms

    810ms

    830ms

    160ms

    1800ms

    1350ms

    900ms

    450ms

    0ms

  • Maximum latency (at 300 calls) in filter extraction problems

    GPT - 4o

    mini

    GPT - 4o

    GPT - 3.5

    turbo

    Claude 3.5

    Sonnet

    Sineps

    1350ms

    3547ms

    9100ms

    4900ms

    382ms

    9000ms

    6750ms

    4500ms

    2250ms

    0ms

Highly Accurate

Highly Accurate

GPT-4o level accuracy

GPT-4o level accuracy

Accuracy in filter extraction problems

GPT - 4o

mini

GPT - 4o

GPT - 3.5

turbo

Claude 3.5

Sonnet

Sineps

69.8%

96.6%

97%

83.6%

99.4%

100%

90%

80%

70%

60%

Accuracy in filter extraction problems

GPT - 4o

mini

GPT - 4o

GPT - 3.5

turbo

Claude 3.5

Sonnet

Sineps

69.8%

96.6%

97%

83.6%

99.4%

100%

90%

80%

70%

60%

Multilingual

Multilingual

Supports English, Chinese, French, Korean, Spanish, Italian, and Japanese

Supports English, Chinese, French, Korean, Spanish, Italian, and Japanese

* We will make the dataset publicly available for evaluation.

* We will make the dataset publicly available for evaluation.

Where to use?

Where to use?

Doc 1

Doc 2

Doc 3

> = 2024-01-01 AND < = 2024-12-31

Filter Extractor

Published

date

“Find articles about the impact of

Large language models on the economy,

published in 2024.”

Seach

Engine

Doc 1

Doc 2

Doc 3

> = 2024-01-01 AND < = 2024-12-31

Filter Extractor

Published

date

“Find articles about the impact of

Large language models on the economy,

published in 2024.”

Seach

Engine

It enables more advanced retrieval capabilities than just semantic or lexical search

It enables more advanced retrieval capabilities than just semantic or lexical search

Sineps

Company : Ceres Technologies, Co., Ltd. | CEO : Inhyuk Na

Call : +82 70-7355-4189 | E-mail : infosec@sineps.io

Registration number : 716-81-02382

Address : 577, Dongdaegu-ro, Dong-gu, Daegu, Republic of Korea

ⓒ 2024 Ceres Technologies, All rights reserved.

Sineps

Company : Ceres Technologies, Co., Ltd.


CEO : Inhyuk Na


Call : +82 70-7355-4189


E-mail : infosec@sineps.io


Registration number : 716-81-02382


Address : 577, Dongdaegu-ro, Dong-gu,


Daegu, Republic of Korea

ⓒ 2024 Ceres Technologies,

All rights reserved.

Privacy Policy

Terms of Use

Sineps

Company : Ceres Technologies, Co., Ltd. | CEO : Inhyuk Na

Call : +82 70-7355-4189 | E-mail : infosec@sineps.io

Registration number : 716-81-02382

Address : 577, Dongdaegu-ro, Dong-gu, Daegu, Republic of Korea

ⓒ 2024 Ceres Technologies, All rights reserved.